Search results for "sapiro" - lists.mailman3.org

[MM3-users] Re: Mailman3 not able to deliver emails sent through list, but moderation notifications

by Mark Sapiro

On 11/26/22 05:09, Daniel Böhmer wrote: > > During startup /var/log/mailman3/mailman.log receives: > >> Cannot connect to SMTP server localhost on port 25 ... > I found this article stating the the SMTP error message > might be misleading: > https://blog.tfiu.de/mailman3-cannot-connect-to-smtp-server-localhost-on-po… > > How to further debug this? > What could be wrong with my Mailman3 instance and how to fix it? The article is relevant and mentions the most likely issue which is lack of permission on one or more message templates. Everything in Mailman's var/ directory should be owned by Mailman's user:group - normally `list:list` in Debian packages. I'm only guessing here, but I suspect you ran `mailman import21` at some point as `root`. Never run `mailman` commands as `root`. Only run them as the Mailman user - `list` in Debian packages. Note that Mailman core in Debian Buster is 3.2.1. Since 3.3.5, "The mailman command will refuse to run as root unless the new --run-as-root option is specified." in order to help avoid this issue. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

2 years, 8 months

[Test] Re: What now -reply from web

by Mark Sapiro

On 03/20/2016 07:22 PM, mailman(a)mailman3.org wrote: > The last post with SSL = True and latest core still hits the same > underlying exception both in trying to get the archive URL and in > trying to archive the message. The core now handles these exceptions > and continues so the message goes to the list, but isn't archived and > has no archive headers. > > This is a post from hyperkitty which will probably be handled the > same, so we have to go back to SSL = False until we can fix the > certificate issue. Yes. Bottom line. With USE_SSL = True, core connects to hyperkitty via SSL to 127.0.0.1 and the certificates I installed don't include 127.0.0.1 in their domains, so urllib3 or requests (depending on where we're coming from) throws SSLError: hostname '127.0.0.1' doesn't match either of 'lists.mailman3.org', 'mirror.list.org', 'mirror.mailman3.org'. The core now handles the exceptions and the message is ultimately delivered but not archived. regardless of where it originated. We're back to USE_SSL = False now so this post should be archived, albeit the archive URLs will have 127.0.0.1 ast the host and not lists.mailman3.org. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

9 years, 4 months

[MM3-users] Re: WEB UI fail to reply and create new thread

by Mark Sapiro

On 12/19/20 3:02 AM, Carlo Landmeter wrote: > We are not running in debug mode, we simply set the loggers to log DEBUG or else we don't see much info. > I also doubt the shown error is the problem as I see more template errors. > I have also tried to run the official docker images mailman-docker but im having the same issue. > > When I post via the online form, i can see uwsgi requests with http status 200 from the ajax call, so i would expect it to work, but it does not leave a message in hyperkitty nor sends an email. > > PS emails are send just fine via smtp posting. The error reported from hyperkitty/ajax/temp_message.html in the OP has nothing to do with the replies not being sent or archived. hyperkitty/templates/hyperkitty/ajax/temp_message.html is called to build the HTML response sent to the web browser after the reply has been sent to Mailman core. Abhilash's fix at <https://gitlab.com/mailman/hyperkitty/-/merge_requests/299> will get rid of that error, but I would be very surprised if it fixes the problem. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

4 years, 7 months

[MM3-users] Re: [Mailman-Users] Re: Member posting to a list but address not in membership list.

by Mark Sapiro

On 1/12/22 7:42 PM, Adam Morris wrote: > Hi Mark, > I've looked in all these places and the address isn't anywhere. (please keep the thread on the list) Then the address that's posting isn't what you think it is or something in your installation is broken. As I said, it could be that Reply-To: or Sender: is the list member, but depending on list settings, these things get overwritten in mail from the list, so you can't really know. If something is broken, you need to discuss this with your hosting provider who, unfortunately, may be no help. Many, not all, cPanel hosts offer mailman because it comes with cPanel, but actually know nothing about Mailman. See https://wiki.list.org/DOC/Mailman%20and%20CPanel > On 13-January-2022 11:31, Mark Sapiro wrote: >> On 1/11/22 7:51 PM, Adam Morris wrote: >>> I have someone who is posting to a list I run. >>> I subscribed them originally but their address doesn't show up in the >>> list of members so I can't change moderation status etc. >>> Where else could I look to find their address? >> >> >> The first check on a post is if any of the address in From:, Reply-To: >> or Sender: headers or the envelope sender is a list member, i.e. in >> the web admin Membership List. If so, that member's moderation is >> applied. If not, the next check is if the From:the *_these_nonmembers >> lists under Privacy options... -> Sender filters and finally Privacy >> options... -> Sender filters -> generic_nonmember_action. >> >> If a post to the list is being accepted without being held for >> moderation, either some address in From:, Reply-To: or Sender: headers >> or the envelope sender is in the Membership List with `mod` unchecked, >> or the From: or possibly Sender: address is in accept_these_nonmembers >> or generic_nonmember_action is Accept. >> > > -- > > Adam Morris > Mobile: 0414 431105 > Email, iMessage & FaceTime > adam(a)damorris.com > -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

3 years, 6 months

[MM3-users] Re: Mixing Mailman 3 with AWS SES

by Mark Sapiro

On 3/24/21 7:58 AM, Brian Carpenter wrote: > On 3/24/21 9:54 AM, Ralph Weber wrote: >> Comments received directly from one of the people who developed the >> AWS Marketplace kits for Mailman 3 and Mailman 4 have definitively >> established that AWS Mailman is *not* intended to interact with AWS >> SES when receiving email messages > > I don't think there is a Mailman 4 and the person who is marketing his > services under that name is sending a confusing message to the public. Danil Smirnov owns both the mailman3.com and mailman4.com domains. Mailman 4 is apparently his branding of "Mailman 3 in AWS Marketplace". In response to our request, the following is included at the bottom of <https://mailman3.com/>: Mailman3.com (this site) is unaffiliated with Mailman3.org, the domain used to coordinate development of the Mailman 3 free software suite. The Mailman 3 core developers are not involved with Mailman3.com, and cannot answer any questions or provide any support related to its operations. If you are looking for resources for development of the Mailman 3 free software suite, please see mailman3.org. Yes, this is all ripe for confusion ... -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

4 years, 4 months

[MM3-users] Re: Upgrading mailman 3 venv installation

by Mark Sapiro

On 9/29/21 10:52 AM, Morris Jones wrote: > > In my virtual environment, I ran `pip list -o` and got a significant > list. I noticed that all of the mailman packages seemed to be "Type > sdist" and the framework packages "Type wheel". (Dunno what "Type" means > in this context.) Type is the package type. sdist is a source distribution. wheel is a form of binary distribution. Older forms are egg and bdist. > So I tried `pip list -o | grep sdist` and got a smaller list: > > (venv) mailman@mail:~$ pip list -o | grep sdist > atpublic 2.1.1 2.3 sdist > django-allauth 0.43.0 0.45.0 sdist > django-mailman3 1.3.4 1.3.7 sdist > flufl.i18n 3.1.3 3.2 sdist > flufl.lock 5.0.3 6.0 sdist > HyperKitty 1.3.3 1.3.4 sdist > lazr.config 2.2.2 2.2.3 sdist > mailman 3.3.2 3.3.5 sdist > mailmanclient 3.3.1 3.3.3 sdist > mysqlclient 2.0.1 2.0.3 sdist > postorius 1.3.3 1.3.6 sdist > > Is this a reasonable list of the packages that I should update? Several of those are dependencies. The packages you should upgrade are mailman, postorius, hyperkitty, django-mailman3 and mailmanclient. Dependencies will be upgraded as required. If a dependency is not upgraded, the older version should work. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

3 years, 10 months

[MM3-users] Re: Multiple server replicas

by Mark Sapiro

On 5/23/23 19:33, Dave Hall via Mailman-users wrote: > > With Mailman 3 we would like to keep the same kind of setup. However, > things have changed quite a bit. Having both mailman and mailman-web > instances all share their respective MySql/MariaDB databases is the easy > part. For /var/lib/mailman3 it is a little less clear. So far I have > symlinked messages, archives, lists, data, and templates onto the NFS > share. I'm wondering about cache, locks, queue, and web/fulltext_index. cache is used to store the contents of templates which are retrieved by URI other than mailman:// URIs. I.e. mailman:// URIs refer to templates in the templates directory or in the Mailman installation itself so are cheap to retrieve. Other URIs can refer to possibly remote servers and may be expensive to retrieve so the contents are cached. In short, this should be shared. locks contains the master lock and others locking things used by this instance. This should not be shared. queue is the queues for this instance and should not be shared. In your installation, web/fulltext_index is the search index for whoosh which is apparently the hyperkitty configured search engine. This should be shared. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

2 years, 2 months

[MM3-users] Re: Mailman3 doesn't send Mails to users (Debian12 with Plesk)

by Mark Sapiro

On 5/24/24 2:18 AM, marcus--- via Mailman-users wrote: > > The problem, however, is that it does not send out any mails, nor do they appear in the archive. > > When I look in the Mailman-Log I find: > > May 16 01:44:55 2024 (6833) Cannot connect to SMTP server localhost on port 25 This is a misleading error message. It can occur when Mailman's out runner can't read a template. You need to check ownership and permissions on mailman/templates in the code tree and Mailman's var/templates directory which probably has permissions errors from running inport21 as root. Also, if you've set any templates in Postorius, they may not be accessable if your setting for POSTORIUS_TEMPLATE_BASE_URL is not good. > May 16 01:48:15 2024 (61252) error opening connection to nntp_host: localhost > [Errno 111] Connection refused > May 16 01:48:15 2024 (61252) NNTP error for list vereinsmitglieder(a)ludothek.de: > [Errno 111] Connection refused Are you gating messages to/from usenet for this list? If so, you have to set host: and port: and maybe user: and password in the nntp section of mailman.cfg -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

1 year, 2 months

[MM3-users] Re: Database permission problem?

by Mark Sapiro

On 10/15/24 15:24, Bill Oliver wrote: > > Yeah, I saw that, too. The double db files are created by postmap. I > haven't worried about it too much yet, since the permissions are the > same. I don't know why postmap does this, but it's next on my list of > things to fix. You should not be running postmap manually and never on a *.db file in any case. Mailman should be creating postfix_domains, postfix_lmtp and postfix_vmap and calling postmap on those to create the .db files. > In my main.cf postfix_lmtp is mentioned four times: > > virtual_mailbox_maps = > proxy:mysql:/etc/postfix/sql/mysql_virtual_mailbox_maps.cf, > > proxy:mysql:/etc/postfix/sql/mysql_virtual_alias_domain_mailbox_maps.cf > #added 20241009 for mailman > hash:/opt/mailman/mm/var/data/postfix_lmtp ${dovecot} > > virtual_alias_maps = > proxy:mysql:/etc/postfix/sql/mysql_virtual_alias_maps.cf, > proxy:mysql:/etc/postfix/sql/mysql_virtual_alias_domain_maps.cf, > > proxy:mysql:/etc/postfix/sql/mysql_virtual_alias_domain_catchall_maps.c > f > #added 20241009 for mailman > hash:/opt/mailman/mm/var/data/postfix_vmap > > #Mailman transport stuff > transport_maps = > hash:/opt/mailman/mm/var/data/postfix_lmtp > local_recipient_maps = > hash:/opt/mailman/mm/var/data/postfix_lmtp ${dovecot} Are you adding ${dovecot} to the local_recipient_maps and virtual_mailbox_maps mailman entries or was that there before? -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

9 months, 2 weeks

[MM3-users] Re: Outgoing messages fro Mailman fails DMARC

by Mark Sapiro

On 04/19/2018 03:22 AM, Henrik Rasmussen wrote: > > Thanks you for the links. According to https://wiki.list.org/DEV/DMARC the MM3 list setting "DMARC mitigation action" (which actually say "Replace From: with list address" in Postorius) should change the From: header. This doesn't seem to work if I don't also set "DMARC Mitigate unconditionally" to Yes, but does if I do. If DMARC Mitigate unconditionally is No, DMARC mitigation action will only be applied if the original From: domain publishes a DMARC policy of quarantine or reject. If the action is not being applied in this case, there is some issue in looking up the DMARC policy for the domain. Possibly, dnspython is not available or there is some other issue with this. In any case, errors should be logged in Mailman's log. > Where do I change defaults for all lists, instead of a per list change in settings/dmarc_mitigations ? You need to do this with list styles. See <http://mailman.readthedocs.io/en/latest/src/mailman/styles/docs/styles.html>. At present, Postorius doesn't allow specifying a style when creating a list, so you may need to modify mailman/src/mailman/styles/base.py. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

7 years, 3 months

[MM3-users] Re: Adding / manipulating users

by Mark Sapiro

On 11/18/19 6:15 AM, Seth Seeger wrote: > > My next question - I have two lists from Yahoo to import. I’ve migrated them through groups.io and have a convenient export. I now have all the data in formats I can handle. (json file for members, mbox for archives.) What’s the best way to import hundreds of users, setting their mail delivery status and moderation status at the same time? (There doesn’t seem to be a command line option for adding members anymore where you can also set their delivery status.) I’ve looked at the documentation, but couldn’t figure out what to do. You are correct that there is no existing command line tool or Postorius way to do this. If you have facility with Python, you can look at the mailman import21 command which is actually processed by mailman/utilities/importer.py. It shouldn't be too difficult to adapt the part that imports the list members to use your data and create a script to do what you want. Alternatively, you could create a script that would post the appropriate transactions to add a member and set moderation and delivery status to Mailman's REST API and use that to process your members. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

5 years, 8 months

[MM3-users] Re: Hyperkitty: Error when rebuilding Index

by Mark Sapiro

On 11/26/21 4:35 AM, M4tze wrote: > Dear list members, > > if there is any Hyperkitty expert here, I have issues in Hyperkitty > (just updated Maxkings Docker images) when rebuilding the Index. > > Unfortunately, there are more than 25,000 mails and I do not know which > one is/might be broken. Does the error mean that the rebuilding process > is incomplete or does it simply skip the erroneous message? It means the process is incomplete. > Traceback (most recent call last): ... > return self._result.fetch_row(size, self._fetch_type) > UnicodeDecodeError: 'utf-8' codec can't decode byte 0xed in position > 3209: invalid continuation byte I'm not sure what's going on here. I suspect the issue is withe the `content` of some message in the hyperkitty_email table in the database. You could try a database query like ``` SELECT * FROM hyperkitty_email WHERE LEN(content) >= 3209; ``` If you're lucky, that may just return a few messages and enable you to find the problem one. Otherwise, you could try first running `mailman-web clear_index` and then run `mailman-web update_index` with --start and --end options to specify a date range and zero in on the offending message. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

3 years, 8 months

[MM3-users] Re: Mailman3-web/Postorius behind Lighttpd

by Mark Sapiro

On 2/16/22 10:00, Élodie BOSSIER via Mailman-users wrote: > > I'm using Mailman 3.3.3 (Tom Sawyer) on Debian Bullseye 11.2 from > packages : When you install Debian packages, your primary support resource is Debian. ... > I can access from localhost to http://127.0.0.1:8000/ which move me > correctly to http://127.0.0.1:8000/postorius/lists/ and all seem to be > correct. > > But 127.0.0.1:8000 listen only in localhost : > #netstat -tanpu | grep 8000 > tcp 0 0 127.0.0.1:8000 0.0.0.0:* LISTEN > 626175/python3 Which is as it should be. > So I try to use my lighttpd as proxy as recommanded with this config : > > >> scgi.protocol = "uwsgi" >> scgi.server = ( >> "/" => (( "host" => "127.0.0.1", "port" => 8000, "check-local" => >> "disable" )), >> ) > > but it's return me a "502 Bad Gateway". I don't know anything about lighttpd, but the fact that you can access http://127.0.0.1:8000/ from localhost and that seems to work suggests that you shouldn't be using uwsgi protocol, but just a straight http proxy. See https://redmine.lighttpd.net/projects/lighttpd/wiki/HowToPythonWSGI#lighttp… and https://redmine.lighttpd.net/projects/lighttpd/wiki/Docs_ModProxy -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

3 years, 5 months

[MM3-users] Re: Fwd: [Django] ERROR (EXTERNAL IP): Internal Server Error: /archives/search

by Mark Sapiro

On 1/28/23 09:39, Odhiambo Washington wrote: > Now that you have mentioned > https://gitlab.com/mailman/mailman-suite-doc/-/blob/master/source/install/v…,, > please show me how to import it, edit. Then I can submit a diff. > Alternatively, I'd love to be given access to edit that document (and the > edit can be reviewed before getting published) to make it easy to follow by > someone with zero python background. > I am willing to invest some time in improving it for laymen. There are multiple ways to proceed. The simplest is to just download https://gitlab.com/mailman/mailman-suite-doc/-/blob/master/source/install/v… (there's a download button on that page), edit that and post the edited file somewhere where we can get it to look at. The better way is more complex. That way is for you to create a personal project space at https://gitlab.com/users/sign_up and clone the mailman/mailman-suite-doc project into your personal space and edit it there and then create a merge request to merge your edits into the mailman/mailman-suite-doc project. GitLab documentation is at https://gitlab.com/help -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

2 years, 6 months

[MM3-users] Re: configuring postfix local recipient table

by Mark Sapiro

On 4/23/25 9:25 AM, Stephen J. Turnbull wrote: > Alexander Inzinger-Zrock via Mailman-users writes: > > and then i just found out that my postfix_lmpt file was empty > > 1. What database are you using?[1] Is it running? If not, yes, that > file is going to be empty. > 2. Do you have any lists configured for your Mailman 3 instance? If > not, that file will be empty. > 3. Is Mailman running? (This probably isn't necessary, just > checking.) > 4. Is there a postfix_lmtp.db file in the same place? (Postfix will > not read the plaintext version, it must be a .db file.) If you haven't set `incoming` in the `mta` section of mailman.cfg to other than the default `mailman.mta.postfix.LMTP`, the `mailman aliases` command will at least write ``` # AUTOMATICALLY GENERATED BY MAILMAN ON {} # # This file is generated by Mailman, and is kept in sync with the binary hash # file. YOU SHOULD NOT MANUALLY EDIT THIS FILE unless you know what you're # doing, and can keep the two files properly in sync. If you screw it up, # you're on your own. ``` to the postfix_lmtp file. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

3 months, 1 week

[MM3-users]Re: Installation issues

by Mark Sapiro

On 04/19/2016 11:40 AM, Andrew Hodgson wrote: > > I have often wondered if there was a howto on how this specific mailman3.org server was built, so it gives us a production-like template to follow? Sort of, but it's not (yet) public. The process I followed is very convoluted and not what I would recommend anyone follow. Part of the issue is I started with mailman-bundler not using a venv, but bundler still installs a lot of stuff in Python 2 and Python 3 venvs. I then proceeded to install the git master branches of mailman, mailmanclient, mailman-hyperkitty, hyperkitty and postorius and their dependencies system wide and ripped out most of the venv stuff. Then there were issues because I'm running django 1.9 only and that's not compatible with django_haystack 2.4.1, so I upgraded that to 2.5.dev0 from GitHub, but that now has a compatibility issue with HyperKitty which I have worked around. As you can see this is not a clean process. If I were starting over with what I've learned, I'd probably let bundler install it all in venvs and then upgrade the venvs to the git master branches. There would still be issues, but it would be cleaner. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

9 years, 3 months

[MM3-users] Re: How to change the language in mailman3

by Mark Sapiro

On 7/17/20 8:22 AM, maszap(a)tin.it wrote: > Hi everyone, > from here: > https://hosted.weblate.org/projects/gnu-mailman/hyperkitty/it/ > I see that there is a way to change the language of mailman3, postorius and hyperkitty > But I don't find how. > I read that you can download the .po files for the chosen language, but I didn't understand how to use these files. > Could anyone show me how to do it? The various message catalogs for Mailman core are in mailman/messages/ and those for hyperkitty, django-mailman3 and postorius are respectively in hyperkitty/locale/, django-mailman3/locale/ and postorius/locale/. The translations from weblate are incorporated in the distribution, but depending on how Mailman is installed, may be out of date. To set language for Django (i.e. hyperkitty and postorius) see <https://docs.djangoproject.com/en/3.0/topics/i18n/translation/#how-django-d…>. For Mailman core, set `[mailman] default_language:` in mailman.cfg. You also need to (re)generate the .mo file which can be done with the generate_mo.sh script. See <https://docs.mailman3.org/projects/mailman/en/latest/src/mailman/docs/inter…> For Django, you need to generate .mo files with `django_admin compilemessages` -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

5 years

[MM3-users] Re: Cloned mailman3 from the gitlab site, having issues locating the mailman3 file to start it

by Mark Sapiro

On 5/17/24 12:06 PM, Keith Christian wrote: > The path to the mailman3.py file is: > /home/mailman3/mailman/src/mailman/bin/mailman.py > I created a mailman3 user, then, set the ownership on all directories > and files in /home/mailman3 and below to mailman3:mailman3. You need to install all the dependencies and also have /home/mailman3/mailman/src/mailman in your PYTHONPATH. If you just want Mailman Core and given that you've already cloned it from gitlab, I suggest creating a Python virtual environment as the mailman3 user with `python3 -m venv /path/to/your/venv` and then as the mailman3 user activate the venv, `source /path/to/your/venv/bin/activate`, and do `pip install /home/mailman3/mailman` to install Mailman and it's dependencies in the venv. Then you can just run `mailman info`. I suggest a venv for this to isolate the dependencies from things that might already be installed system wide with differing versions. To install the whole suite including the Postorius web UI for list management and the HyperKitty archiver, see https://docs.mailman3.org/en/latest/install/virtualenv.html -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

1 year, 2 months

[MM3-users] Re: Separating Lists by Domain

by Mark Sapiro

On 1/3/19 2:34 PM, Abhilash Raj wrote: > > > On Thu, Jan 3, 2019, at 2:27 PM, brian(a)emwd.com wrote: >> So who manages Postorius? Is that a whole separate application from >> Mailman 3? Is there a way we can get whoever the developer is to update >> it where we can separate lists by domain the way Hyperkitty does? Mark >> Sapiro refers to this in the following post: >> >> https://lists.mailman3.org/archives/list/mailman-users@mailman3.org/message… > > You can file an issue at https://gitlab.com/mailman/postorius/issues Yes, filing a GitLab issue is the best way to get things on the record. It's no guarantee that things will be addressed immediately, but they are there, and merge/pull requests are always welcome. As far as your question goes, Mailman Core, Postorius and HyperKitty are separate entities and they were mostly developed by separate people/groups and some important players have either overtly or implicitly retired from the project. Not long ago I write in another context > MM3 is much more complex than MM2.1 and there's a lot that I don't fully > understand myself. > > There is a basic core engine which does all the work of > creating/managing lists and users and receiving and sending mail. It has > no web UI per se. It exposes a RESTful API on the host server. > > The web management UI is Postorius and it is based on the Django web > framework and communicates with core via the REST API. The Archiver is > HyperKitty and it is fairly independent. It is also based on the Django > web framework. Account creation and web login is via django-allauth. > > Postorius, HyperKitty and the core were all developed separately by > different people and no one understands all of it. That said, Abhilash is currently the lead developer for all of Mailman 3 and he's doing a great job. I and a few others are also involved (I'd be more involved if I didn't have Mailman 2.1 to distract me), but we are a small group and things take time. And as I said, merge/pull requests for GitLab issues are always welcome. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

6 years, 6 months

[MM3-users] Re: mailman3 auto-disabling of subscriptions

by Mark Sapiro

On 6/16/23 22:14, Stephen J. Turnbull wrote: > > @Mark > I've been seeing an unusual number of disabled subscriptions on > python.org and mailman3.org lists recently. Maybe we should look into > this. If you haven't seen this O365 breakage, I can send you some > samples. I haven't noticed this. I do receive notices when delivery is disabled, and I generally look at the attached DSN, and I don't recall any that have this issue. We do have an issue with mailman-users(a)mailman3.org messages for five users in the outlook.com, msn.com and hotmail.es domains. These are all since the server was moved last month and are all for the reason 550 5.7.1 Unfortunately, messages from [165.227.190.41] weren't sent. Please contact your Internet service provider since part of their network is on our block list (S3140). You can also refer your provider to http://mail.live.com/mail/troubleshooting.aspx#errors. [VI1EUR05FT020.eop-eur05.prod.protection.outlook.com 2023-06-17T05:14:54.401Z 08DB6E025E06C828] with differences only in the bracketed part at the end. The same five addresses keep getting disabled because I keep re-enabling them, and I have contacted Microsoft about this, so far with no results. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

2 years, 1 month

[MM3-users] Re: Timeout error on anonymous subscribe

by Mark Sapiro

On 10/14/24 06:13, Gerald Vogt wrote: > Hi, > > sorry to bother, but does anyone have an idea why this doesn't work? I > am on mailman 3.3.10 by now but still not difference. I wonder why > nobody else has the same problem. I didn't think my setup would be so > special... > > Thanks, > > Gerald > > vogt＠spamcop.net wrote: >> >> Hi, >> >> I am running a mailman 3.3.9 server (pip installation). It's working >> fine so far. However, I have problems setting a template. I set a >> '[list:user:action:subscribe]' template for a list with some simple >> text like "test1234". How are you creating the template? If you are creating it via Postorius, what is your setting for POSTORIUS_TEMPLATE_BASE_URL and what does the entry in the template table in the database look like? POSTORIUS_TEMPLATE_BASE_URL should be the same URL you use to access Postorius up to but not including the `/postorius/` or `/mailman3/` part. I.e, for this list whose info page is at https://lists.mailman3.org/mailman3/lists/mailman-users.mailman3.org/, the setting is ``` POSTORIUS_TEMPLATE_BASE_URL = 'https://lists.mailman3.org' ``` -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

9 months, 1 week

[MM3-users] Re: MTA setup

by Mark Sapiro

On 11/17/24 19:32, Arte Chambers via Mailman-users wrote: > > Mailman is successfully sending invite emails however user messages sent to > the list are not being sent to subscribers. Also in Hyperkitty none of my > test messages to the list are showing up in the archive. Are there messages in Mailman's shunt queue? Check for errors in mailman.log. Are there messages for hyperkitty queued in Mailman's var/archives/hyperkitty/spool/ directory? > I am using Postfix with an SMTP relay hosted at Brevo.com (formally > sendinblue). Brevo's logs show the message getting rejected > > > Log from brevo.com (SMPT Relay) > Sent on 17-11-2024 21:49:16 >>From listmember(a)gmail.com > To listmember(a)gmail.com > Message ID<CAPesOD1y0fEaG=GFodz=Lgs1m_hqRErPnQcpXCNtKsBbJ= > BJ7Q(a)mail.gmail.com> > > 17-11-2024 21:49:16 > Sending has been rejected because the sender you used listmember(a)gmail.com > is not valid. Validate your sender or authenticate your domain This appears to be something you need to do for the MTA at smtp-relay.sendinblue.com, or possibly something more is needed in /etc/postfix/sasl_passwd. > #SMPT Relay Brevo > relayhost = [smtp-relay.sendinblue.com]:587 > smtp_sasl_auth_enable = yes > smtp_sasl_security_options = noanonymous > smtp_sasl_password_maps = hash:/etc/postfix/sasl_passwd -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

8 months, 2 weeks

[MM3-users] Re: Confirmation for subscriptions

by Mark Sapiro

On 6/23/19 7:24 AM, andrew.bernard(a)gmail.com wrote: > Suddenly stopped working. For all the lists I have created, suddenly the confirm setting is ignored for subscriptions and users are subscribed right away. I assume you are looking at the Postorius List Settings -> Subscription Policy. That should reflect the setting in core. How are these users requesting subscription/being subscribed? > This only happened after I added another domain - surely that can't be the cause? Unlikely, but not impossible. > Where to look to troubleshoot this? You could use `mailman shell` to examine the setting in Mailman core, but it's unlikely that would be different from what Postorius shows. Other than that, you can look in logs, but there is not very much logged for the subscription workflow. To be more help, I'd like to know exactly when this happens. First, for Postorius Mass subscribe, this is expected. Other than that, does it happen on email to the list-join address? Does it happen with subscription via the form on the Postorius list Info page? Is it possible the user has a 'vacation' program that is responding to the confirmation request? See <https://gitlab.com/mailman/mailman/issues/23> (we should fix that) -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

6 years, 1 month

[MM3-users] Re: How to identify the exact reason behind the shunting of the mails ?

by Mark Sapiro

On 6/18/21 6:48 AM, Shashikanth Komandoor wrote: > Dear Mark, > > FYI, I observed this issue with the postfix that receives the > mails from the previous MTA and hands it over to the mailman. But I fixed > at postfix by upgrading it to the version postfix 3. > > But still Mailman is not able to understand Hindi (Indian local > language) in which language the display names of the email ids are there. > > So, now what to do at Mailman so that it can also understand > local languages or any work around or other alternate solutions for these > sort of cases. As noted in my reply at https://lists.mailman3.org/archives/list/mailman-users@mailman3.org/message… the issue is a Mailman bug which is fixed in the next release. It has to do with the fact that the display names are RFC 2047 encoded and not the particular encoding. Actually, the fix at https://gitlab.com/mailman/mailman/-/merge_requests/841 depends on https://gitlab.com/mailman/mailman/-/merge_requests/652 which was released in Mailman 3.3.2. A complete patch for versions < 3.3.2 is attached. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

4 years, 1 month

[MM3-users] Re: Postorius no connection to REST API

by Mark Sapiro

On 8/13/21 7:18 AM, Richard Rosner wrote: > > I set them all to https. which solved the https error. But I still get Error 500. >>From apache log: > [Fri Aug 13 15:47:58.833305 2021] [mpm_event:notice] [pid 26720:tid 139986206332032] AH00489: Apache/2.4.48 (Debian) OpenSSL/1.1.1d mod_wsgi/4.6.5 \ > Python/3.7 configured -- resuming normal operations > [Fri Aug 13 15:47:58.833436 2021] [core:notice] [pid 26720:tid 139986206332032] AH00094: Command line: '/usr/sbin/apache2' > [Fri Aug 13 15:49:41.673108 2021] [ssl:error] [pid 26722:tid 139985899259648] [remote 127.0.0.1:8000] AH01961: SSL Proxy requested for lists.fsmuw.\ > rwth-aachen.de:443 but not enabled [Hint: SSLProxyEngine] > [Fri Aug 13 15:49:41.905868 2021] [proxy:error] [pid 26722:tid 139985899259648] AH00961: https: failed to enable ssl support for 127.0.0.1:8000 (12\ > 7.0.0.1) > [Fri Aug 13 15:59:25.049708 2021] [ssl:error] [pid 26722:tid 139985899259648] [remote 127.0.0.1:8000] AH01961: SSL Proxy requested for lists.fsmuw.\ > rwth-aachen.de:443 but not enabled [Hint: SSLProxyEngine] > [Fri Aug 13 15:59:25.059217 2021] [proxy:error] [pid 26722:tid 139985899259648] AH00961: https: failed to enable ssl support for 127.0.0.1:8000 (12\ > 7.0.0.1) You need SSLProxyEngine on and possibly one or more of the SSLProxyMachineCertificate* directives in your apache config to proxy via https. See <https://httpd.apache.org/docs/2.4/mod/mod_ssl.html#sslproxymachinecertifica…> et seq. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

3 years, 11 months

[MM3-users] Re: Problem creating list members using REST API

by Mark Sapiro

On 1/18/22 10:53 AM, Stephen Daniel wrote: > 2nd follow-up. Apologies for spamming this list. > > I went wandering through the code as suggested by the traceback. I > discovered that thee "as_boolean" routine is expecting a string argument > but getting a boolean argument. > > So, If I send the booleans as strings, with values "true" or "false", all > is well. > This is not what the API documentation suggests I should do. Can someone > either fix the documentation or the code? Those snippets in the documentation are doctests. That code is actually executed by unit testing and passes. In all those snippets, dump_json is passing the second argument as `data` to mailman.testing.documentation.dump_json() which in turn calls mailman.testing.documentation.call_http() which calls mailman.testing.helpers.call_api() which ultimately calls requests.request() with the same `data` argument. So basically, the issue is the difference between how requests.request() handles that `data` dictionary in creating the POST data and whatever you are doing. Possibly we could add some caveat to <https://docs.mailman3.org/projects/mailman/en/latest/src/mailman/docs/docum…> or possibly elsewhere. Do you have a suggestion as to what? -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

3 years, 6 months

[MM3-users] Re: Weird startup problem after upgrade

by Mark Sapiro

On 12/2/22 08:21, Odhiambo Washington wrote: > I am facing an odd situation in which MM3 doesn't start successfully on > reboot, but I can start it manually. > This is a venv install. > This started after I did an upgrade (pip install --upgrade mailman > postorius hyperkitty mailman-hyperkitty) followed by the requisite > collectstatic|compress|compilemessages|migrate > post-update commands. Was the venv activated when you ran pip? You should also include django-mailman3 and mailmanclient in the packages to be upgraded although they are dependencies and "should" upgrade anyway. Also, at least one person has noted that --force-reinstall was required to get all dependencies upgraded <https://lists.mailman3.org/archives/list/mailman-users@mailman3.org/message…> > Secondly, it appears that the upgrade did NOT change the versions for the > upgraded components - postorius=1.3.6, hyperkitty=1.3.5, > Mailman Core Version=3.3.5. I was expecting the version to all bump up to > the latest. What does `pip freeze` in the venv show for versions? > Thanks in advance for any insights into what could be causing this. How do you start it manually? `mailman start` or `systemctl start mailman3`? What's the content of /etc/systemd/system/mailman3.service? -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

2 years, 7 months

[MM3-users] Re: default language ignored

by Mark Sapiro

On 10/02/2017 05:46 AM, Marc Sluiter wrote: > Hi, > > one last question for now: how do switch the language of the mails sent by mailman? > I have this in my configuration, but without effect: > > [language.de] charset: iso-8859-1 > [language.de] description: German > [language.de] enabled: yes > > [mailman] default_language: de It has no effect because there are as yet no translations in Mailman 3. You need to create templates in the desired language. See <http://mailman.readthedocs.io/en/latest/src/mailman/rest/docs/templates.html> That refers to setting URLs for templates which does work, but the easiest way is to put them in the default search path in Mailman's file hierarchy at $var-dir/templates/. Sitewide ones in $var-dir/templates/site/, Domain specific in $var-dir/templates/domains/ and list specific ones in $var-dir/templates/lists/. For example, a list specific, English language regular (non-digest) message footer might go in $var-dir/templates/lists/mymaillist.example.com/en/list:member:regular:footer > Is it also possible to change the language for a single list? Yes, although that is not yet exposed in Postorius. The list has a preferred_language attribute which you can set via REST or 'mailman withlist' aka 'mailman shell'. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

7 years, 9 months

[MM3-users] Re: list of subscribers

by Mark Sapiro

On 02/04/2018 11:37 AM (UTC -0800), Peter Münster wrote: > On Sun, Feb 04 2018, Simon Hanna wrote: > >> IIRC one person(postorius account) can also signup at different lists using >> different names (regardless of address). > > You mean the "display_name" of mailmanclient here? I won't use it. > For my usage, I prefer a clear association of email, first name and > last name in one account, no need for "display_name" here. This would seem to be specific to your use case. I think in general what we would want to display is the display_name from the list's member record. > Off topic: I have "Receive own postings = Yes", but I don't receive > copies of my postings to this list... We tried to send your copy of your post to you, but your MTA called it spam. Feb 4 19:38:09 mailman postfix/smtp[29085]: 89D4C8031E: to=<pmlists(a)free.fr>, relay=mx1.free.fr[212.27.48.7]:25, delay=1.3, delays=0.04/0/0.81/0.41, dsn=5.0.0, status=bounced (host mx1.free.fr[212.27.48.7] said: 550 spam detected (in reply to end of DATA command)) (time stamp is UTC -0000) I suspect it doesn't like the fact that it's From: you but not sent by your server. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

7 years, 5 months

[MM3-users] Re: Weird hourly indexing error

by Mark Sapiro

On 5/5/20 8:55 AM, Mark Dadgar wrote: > I moved my mail server to a Digital Ocean droplet. I basically installed all the appropriate packages and then rsync’d everything that matters from my metal server up to the cloud instance. Everything more or less works fine. > > However, I am now getting this hourly error - see below. ... > [ERROR/MainProcess] Failed indexing 1 - 1000 (retry 5/5): Term too long (> 245): XSUBJECT95251413 (pid 36301): Term too long (> 245): XSUBJECT95251413 > Traceback (most recent call last): > File "/usr/lib/python3/dist-packages/haystack/management/commands/update_index.py", line 97, in do_update > backend.update(index, current_qs, commit=commit) > File "/usr/local/lib/python3.8/dist-packages/xapian_backend.py", line 488, in update > database.replace_document(document_id, document) > xapian.InvalidArgumentError: Term too long (> 245): XSUBJECT95251413 This error comes from Xapian. It is due to a (new) message in your archive that contains a 'word' longer than 245 characters. There is a PR at <https://github.com/notanumber/xapian-haystack/pull/181> that addresses this plus another suggested fix in that PR's comment thread. Note: I got there by googling "xapian term too long" which gave <https://github.com/notanumber/xapian-haystack/issues/153> as the first hit and that led me to the above. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

5 years, 2 months

[MM3-users] Re: File "/usr/lib/python2.7/httplib.py", line 1148, in getresponse raise ResponseNotReady()

by Mark Sapiro

On 9/28/20 12:58 AM, r.woithe(a)callassoftware.com wrote: > Hello, > i set the necessary config params like descriped at https://wiki.list.org/x/17892072 with timeout = 9000. > But we still get the nigthly email. If you are still using Mailman core 3.2.2, those settings have no effect because the REST server doesn't use gunicorn prior to 3.3.0. See <https://lists.mailman3.org/archives/list/mailman-users@mailman3.org/message… > I guess the best way would to upgrade to mailman-core 3.3.3 to avoid this error. > Unfortunatly i can not use the ubuntu package-manager... > Or is there a second way to avoid this error? I don't know if upgrading Mailman core would avoid the error or not. As I indicated, it's almost certainly not a REST timeout after all. We need to understand what causes the error. As I've said, the sync_mailman goes through the message senders and tries to set the senders mailman_id. If you apply the attached patch to hyperkitty/models/sender.py, it will log the error and ignore it. Then, we'll get a bit more info from the log message. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

4 years, 10 months

[MM3-users] Re: Extracting mailing list users

by Mark Sapiro

On 1/2/23 07:06, Jan Eden via Mailman-users wrote: > On 2023-01-02 14:36, Eggert Ehmke via Mailman-users wrote: > >> Thank you, that worked! Now all my addresses are clean. >> >> Still, don't you think this should be possible via Postorius, or by some mailman cli command, like >> mailman rename_member <old> <new> Yes, it should be possible to change the display name for an address via Postorius. See https://gitlab.com/mailman/postorius/-/issues/451. >> It happened I had still the patched cli_members.py where I had both the old and new line active: >> print(formataddr((address.display_name, address.original_email)), file=outfp) >> print(f'{address.display_name} <{address.original_email}>', file=outfp) >> The first (old) line still produces the output with utf-8 codes. Is this to be expected? > > I would assume the original line outputs the QP-escaped names, to be > used in an email header. Yes. The original `formataddr((address.display_name, address.original_email))` creates the format `display name <email>` suitable for use in an email header. Thus, if the display name contains non-ascii, it is RFC 2047 encoded. The encoding may be quoted-printable or base64 depending on which is shorter. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

2 years, 6 months

[MM3-users] Re: Social signup?

by Mark Sapiro

On 3/12/23 14:32, Duane Raymond wrote: > > Thanks for clarifying. Once I implemented social login, the buttons for > social login did appear fine on the login page, but not on the signup > page. I noticed for this list as well, there are no social signup > buttons. The buttons do not appear on the sign-up page. They probably should, but they don't. > So if it works, how does this happen? > > 1. Is there something I am missing adding to my settings.py and > settings_local.py files to make it appear on the Posterius/Hyperkitty > signup (vs login) page? (I searched but couldn't find anything) You aren't missing anything. > 2. Do I manually add this code to the Posterius/Hyperkitty signup > template? No > 3. If a user tries to login via the login page and isn't registered will > it be counted as a signup? Yes, that's how it works. > I'm trying to get my head around where this is working from / what I need > to do to get social signup obvious on the signup page (or other route) I just created https://gitlab.com/mailman/django-mailman3/-/issues/67 for this. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

2 years, 4 months

[MM3-users] Re: Emergency moderation and clearing out a lot of held messages

by Mark Sapiro

On 11/11/22 16:36, Caballero, Danny (Dan) wrote: > Thanks for some clarification. > > I cleared out 2800 held messages via Python shell "mailman -l" option and it took about 28 hours. Anything done via Mailman shell is a connection from Mailman core to the database, correct? That's correct. > Although I can't replicate that many messages via Postorius/DJango, I do see similar slowness there as well at a message per minute or more of processing. > > I recreated the above Mailman shell clean up via Docker container running MariaDB and noticed similar processing time for discarding held messages. That leads me to conclude that the issue isn't network related or specific to AWS RDS. I'm assuming the docker container installation also had all the lists/users? It seems this must be related to a large number of lists/users and we must be doing a number of queries that doesn't scale well. This seems excessive. We'll have to look at the processing to see what we're doing. There have been some recent improvements in scaling. What Mailman core version is this? -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

2 years, 8 months

[MM3-users] Re: Mailing list resends same email multiple times

by Mark Sapiro

On 3/9/23 04:18, 朱超 wrote: > Hello, I need your help. There have question about Scenes: > > Add mailbox a and mailbox b to mailing lists c and mailing lists d, and send mail to mailing list c and mailing list d through mailbox b , among which a has received many mails sent by b, check the log and find: The file about path:/opt/mailman /var/queue/bounces has the content of retrying mail, but I turned off process bounces. > > Mailing list sends mail via exim, logs found in exim： 023-03-09 00:54:47 1pa4YR-001xk5-0y <= <> R=1pa4YM-001xje-1F U=exim P=local S=7672 T="Mail delivery failed: returning message to sender" from <> I need more information. Even though the list isn't brocessing bounces, bounces will still be returned to the -bounces address and queued in /opt/mailman/var/queue/bounces. If they remain there, the bounce runner is not processing them. If the list's bounce processing is turned off, the bounce runner should just dequeue the message and discard it. If multiple messages go to the list, something is returning them to the list posting address instead of the list-bounces address. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

2 years, 4 months

[MM3-users] Re: Mailman3 on FreeBSD

by Mark Sapiro

On 08/12/2017 04:51 PM, Barry Warsaw wrote: > On Aug 10, 2017, at 16:32, John Poltorak <karotlopj(a)gmail.com> wrote: >> >> On FreeBSD they ought to be in /usr/local/etc/mailman.cfg however I don't >> see any indication as to what I should put in this file and if it was in >> /usr/local/etc/mailman.cfg then how would mailman find it? > > It wouldn’t by default. Right now Mailman does not search for /usr/local/etc/mailman.cfg, but that wouldn’t be a difficult change to add. I’m not sure whether that would makes sense as a general default or not. I’m not up on the FreeBSD conventions and standards here; can you explain why /usr/local/etc is preferred over /etc on that distro? I was hoping Abhilash would chime in here, because I think John is using the docker image and there's probably a specific path for mailman.cfg. As far as what goes in it, it could even be empty if you want to accept all the defaults, but you probably want at least [mailman] site_owner: real_person(a)example.com I expect if you do find -name mailman.cfg in your installation, you'll find at least examples. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

7 years, 11 months

[MM3-users] Re: mailman3 admin from mailman2 & File location questions

by Mark Sapiro

On 11/3/21 1:38 PM, bob B wrote: > Question1 > We use help users with list settings, so we would log into mailman2 with the master password. So what is the best way to duplicate this for mailman3. I assume I can just use the admin account and share that password with my other admins. But I see the domain owners, but is that working? I added someone as a domain owner, and they do not see private email lists and it looks like that can not edit any list settings? There was a post that this had not yet been implemented? Any other options? Domain owners should be able to manage lists in their domains, but currently they can't. You can go to the Django admin interface and give any user superuser status, and then they will be 'admins'. You don't have to share passwords. > Question2 > Where are all the emails/archives stored? are they in /opt/mailman/...., or are they in the database? It depends on the archiver. For the prototype archiver they are in the archives/prototype/ directory in Mailman's var directory. For HyperKitty they are in the database. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

3 years, 8 months

[MM3-users] Re: Deleting messages from HyperKitty archive

by Mark Sapiro

On 3/25/24 12:42, Mark wrote: > Hi, > > I'm running Mailman 3.3.9 with HyperKitty version 1.3.8 (venv installed > as per the docs). > > As superuser (listmaster) I can "Delete message/thread" and "Delete > archive". > > These options aren't there for the list owners. > > The docs (link below) say that in the newer HyperKitty version 1.3.9, > list owners can delete messages/threads > > Is there a patch I can implement on HyperKitty version 1.3.8 (so list > owners can delete messages), or is it possible to update HyperKitty to > version 1.3.9 without bringing everything crashing down? > > *** > https://docs.mailman3.org/projects/hyperkitty/en/latest/news.html > 1.3.9 (2024-02-27) > List owners can now delete messages and threads and reattach threads. > (Fixes #261) > *** The patch is at <https://gitlab.com/mailman/hyperkitty/-/merge_requests/584/diffs>. Upgrading HyperKitty can be done in your venv by something like ``` pip install hyperkitty==1.3.9 ``` however, there may be things that also depend on later versions of django-mailman3, mailmanclient, mailman-web or even mailman core, thus it is best to upgrade everything. For the recommended virtualenv install, see <https://docs.mailman3.org/en/latest/upgrade-guide.html#virtualenv-install> -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

1 year, 4 months

[MM3-users] Re: Data too long for column 'data' in workflowstate

by Mark Sapiro

On 12/14/20 9:46 AM, Morris Jones wrote: > I ran into this error attempting to add a user to a mailing list: > > (pymysql.err.DataError) (1406, "Data too long for column 'data' at row 1") > [SQL: INSERT INTO workflowstate (token, step, data) VALUES (%(token)s, > %(step)s, %(data)s)] > [parameters: {'token': '7394fbf654b3faffc6a65eb95e0616eea6651b9e', > 'step': 'do_confirm_verify', 'data': '{"pre_approved": false, > "pre_confirmed": false, "pre_verified": false, "invitation": false, > "address_key": "dorothyxxxxxxxx(a)gmail.com", "subscriber_key": 1, > "user_key": "00201c01dcc443c3a0d8f6d1ea42e230", "token_owner_key": 1, > "send_welcome_message": null}'}] > > In MySQL I see that column 'data' in workflowstate is VARCHAR(255), and > the data in the query is 259 characters. > > If I have a moment I'll write a bug or patch ... but I'm not really > current on schema migrations. This is <https://gitlab.com/mailman/mailman/-/issues/793> fixed by <https://gitlab.com/mailman/mailman/-/merge_requests/730> You may be able to apply the fix from the merge request, but that depends on your install's latest Alembic Revision ID being 9735f5e5dbdb. Alternatively, I think it may be sufficient to just ALTER TABLE workflowstate MODIFY `data` VARCHAR(510); -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

4 years, 7 months

[MM3-users] Re: mailman import21 passwords

by Mark Sapiro

On 9/16/24 16:25, Sam Darwin via Mailman-users wrote: > > Why can't passwords be imported from mailman2 to mailman3? How difficult would it be to do? > (is there a link to where this is already explained? ) Mailman 3 users have a _password attribute which isn't actually used for anything. The importer used to set this from the MM 2.1 member's password, but this was dropped by <https://gitlab.com/mailman/mailman/-/merge_requests/565>. See <https://mail.python.org/archives/list/mailman-developers@python.org/thread/…> for a discussion of this. Also note that since member passwords were stored in plain text in MM 2.1 and periodically emailed, those passwords are inherently not secure and have possibly been exposed so setting that password, even though encrypted in MM 3, is not a good idea. As far as creating a Django user for an imported member, whether or not with the member's MM 2.1 password, it is at least as expensive as just encrypting the password and would result in poor performance. Further, and perhaps more importantly, Mailman 3 is modular. There is no guarantee that any Mailman Core installation into which a MM 2.1 list is imported actually also supports Django users for Postorius and/or HyperKitty. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

10 months, 2 weeks

[MM3-users] Re: Mail configuration

by Mark Sapiro

On 03/27/2018 03:43 PM, Stephen J. Turnbull wrote: > > My feeling is that you want to ask on Postfix channels, since I don't > see a good reason to involve Mailman itself in what is really a task > for the MTA. Mailman was deliberately designed to not know anything > about outgoing mail except a host and port to send to. Which brings up the point I've been meaning to post for days but it keeps slipping off the back burner and falling on the floor. I can think of a couple of ways to hack the code in Mailman 2.1 to select a different outgoing MTA by list. It can also be done in Mailman 3 by having separate MTA configs per list, but I can't give you a recipe off hand. But, the way I would do this is what you are presumably doing anyway. That is have separate Mailman installations on the old and new server. Then you can move the lists incrementally and gradually. Then all you need is a transport map in Postfix on the old server to route mail to the new server for the already moved lists, and then finally update DNS after all the lists are moved. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

7 years, 4 months

[MM3-users] Re: duplicate dkim headers w/ AWS SES

by Mark Sapiro

On 6/2/20 9:44 AM, sunil soprey wrote: > Periodically, I've seen these errors from few senders, and wondering if community has a idea? Mailman - 3.3.1rc1 Tom Sawyer > > "554 Transaction failed: Duplicate header 'DKIM-Signature'" while sending AWS-SES as the MTA? > > Possible to strip from postfix? I've submitted a AWS ticket. It appears that AWS SES is intended only for delivery of locally generated mail and not mail that is received and resent by a mailing list or forwarding arrangement. This page <https://docs.aws.amazon.com/ses/latest/DeveloperGuide/header-fields.html> gives a list of a large number of headers that can appear only once in a message, among them are Delivered-To:, DKIM-Signature: and a few others that might legitimately appear more than once in messages from a mailing list. There are a couple of ways to remove incoming DKIM sigs. If you are signing with opendkim, there is a RemoveOldSignatures config setting which if true will remove existing signatures when signing. In Mailman itself you can put [mta] remove_dkim_headers: yes in mailman.cfg and Mailman will remove all DomainKey-Signature, DKIM-Signature and Authentication-Results headers from the message before delivery to the MTA. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

5 years, 1 month

[MM3-users] Re: New Mailman 3 installation issues

by Mark Sapiro

On 5/19/21 11:50 AM, Sarah H wrote: > Hi Mark, > > I wanted to make another note on the documentation, the migration documents (https://docs.mailman3.org/en/latest/migration.html#fn2) reference using > > 'python manage.py hyperkitty_import' > > it should be > > 'mailman-web hyperkitty_import' That is a can of worms. It all depends on how you installed Mailman 3. If you follow the doc at https://docs.mailman3.org/en/latest/install/virtualenv.html, the django management command will be `mailman-web`, but installation in other ways can result in its being `manage.py` or `django-admin` or almost anything. For example, https://docs.mailman3.org/en/latest/pre-installation-guide.html#what-do-i-n… refers to a typical Django project in which the management command is `manage.py`. https://docs.mailman3.org/en/latest/config-web.html is inconsistent in itself in that it uses `mailman-web` in most places except the note about qcluster at https://docs.mailman3.org/en/latest/config-web.html#running-the-task-queue which also links to http://hyperkitty.readthedocs.io/en/latest/install.html#asynchronous-tasks which uses `django-admin`. I'm thinking that we should put a note about this near the top of https://docs.mailman3.org/en/latest/ and let it go at that. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

4 years, 2 months

[MM3-users] Re: Block members from sending cc's and bcc's

by Mark Sapiro

On 7/18/25 5:37 PM, Cathryn McGuire wrote: > > The .* pattern did not do the job of blocking cc's and bcc's in the Header Filters of Affinity. It lets them go through, even as hold for moderation has been chosen. First, you can't filter on Bcc because Bcc is removed from the message before being sent. That's the whole point of Bcc. In otherwords, if I mail To: the list with Bcc: to others, the list never sees that Bcc: header and can't filter on it. As far as Cc is concerned, if in Header filters you have a rule with 'cc' in the header box and '.*' in the pattern box, the action box action should be applied to any message with a Cc: header. > I've asked EMWD to advise me because I think Affinity is very new, even though Brian Carpenter was developing it in 2020, is that correct? Affinity is not new, but it is a proprietary product of EMWD. If you have set the header filter as above and it is not working, only EMWD can help with that. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

2 weeks

[MM3-users] Re: old usernames appearing in mailman.log

by Mark Sapiro

On 6/26/23 10:12 PM, Ken Alker wrote: > --On Monday, June 26, 2023 5:03 PM -0700 Mark Sapiro <mark(a)msapiro.net> > wrote: >> >> Your urls.py file probably contains in its urlpatterns list >> ``` >> re_path(r'^mailman3/', include('postorius.urls')), >> re_path(r'^archives/', include('hyperkitty.urls')), >> ``` >> The simplest fix is to add >> ``` >> re_path(r'^postorius/', include('postorius.urls')), >> re_path(r'^hyperkitty/', include('hyperkitty.urls')), >> ``` >> to that list. > > I see many ursl.py files all over the system, but none that appear to be > where I'd expect them to be for processing by any executables. Is this > a file I should have copied over and edited, or created from scratch and > somehow missed? I don't see anything about it in the instructions at > <https://docs.mailman3.org/en/latest/install/virtualenv.html> (although > I do see a post from last July from someone stating it got missed in the > instructions back then). In your case, it is the one in /opt/mailman/venv/lib/python3.x/site-packages/mailman_web/, but it is not a good idea to change that as changes will be overwritten when mailman_web is updated. It's OK to not change it. Just do the following. >> Another way to fix this is in your mailman-hyperkitty.cfg, you have >> ``` >> base_url: https://lists.netlojix.com/hyperkitty >> ``` >> Change that to >> ``` >> base_url: https://lists.netlojix.com/archives >> ``` > > Done, and restarted both services, but I'm not seeing a difference when > I go to <https://lists.netlojix.com/hyperkitty/api/> (I was getting a > "Page not found", and still get that). That's because going to the 'hyperkitty' url won't work, but nothing should be doing that so it's OK if it doesn't work. Fixing mailman-hyperkitty.cfg will tell mailman-hyperkitty to use the 'archives' url, so now it should be working. The next post after you fixed it should archive OK and all the messages in /opt/mailman/mm/var/archives/hyperkitty/spool/ should also be archived and the files removed. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

2 years, 1 month

[MM3-users] Re: Is it possible to add non-members in regex format in Mailman 3 ?

by Mark Sapiro

On 6/16/20 7:09 AM, Shashikanth Komandoor wrote: > > But in mailman 3, I am not allowed to add the non members in > the regex format. Of course it is allowing me to add individual mail ids > and give permissions. But I want to add a non-member which may be something > like the entire cisco.com domain on which I want to set the rule. You cannot add a non-member with a regexp email, but the legacy *_these_nonmembers list attributes still exist in Mailman core and accept regexps. These settings are exposed in Postorius 1.3.3+ under Settings -> Message Acceptance. With older Postorius, you ned to set these directly in core, e.g. with `mailman shell` as in $ bin/mailman shell -l list.example.com Welcome to the GNU Mailman shell Use commit() to commit changes. Use abort() to discard changes since the last commit. Exit with ctrl+D does an implicit commit() but exit() does not. The variable 'm' is the list.example.com mailing list >>> m.accept_these_nonmembers [] >>> m.accept_these_nonmembers = ['^.*[.(a)]good\.domain$'] >>> m.accept_these_nonmembers ['^.*[.(a)]good\\.domain$'] >>> -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

5 years, 1 month

[MM3-users] Re: Membership list changes

by Mark Sapiro

On 2/18/21 9:37 AM, Vako Nicolian wrote: > Using MM 2.1.33 > When I send an email to /{list}-request(a){domain.com} /with /who > <password> /on the subject line, I get a list of the members of the list. > > Here are my questions: > A) Is there a way to export the list to cvs or txt format? copy/paste > B) Can we format the list to have it come in name 1st and then email> C) display number of members > D) Sort the list These things can be done by manipulating the copy/paste result to turn it into a .csv and importing that into a spreadsheet. For example, piping this input ``` Non-digest (regular) members: user1(a)example.com (User One) user2(a)example.com Digest members: user3(a)example.com (Three, User) user4(a)example.com ``` through this script ``` #! /usr/bin/env python import sys from email.utils import parseaddr flag = '' for line in sys.stdin.readlines(): if line.startswith('Non-digest'): flag = 'Reg' continue if line.startswith('Digest'): flag = 'Dig' continue name, addr = parseaddr(line) print('"{}",{},{}'.format(name, addr, flag)) ``` produces this output ``` "User One",user1(a)example.com,Reg "",user2(a)example.com,Reg "Three, User",user3(a)example.com,Dig "",user4(a)example.com,Dig ``` -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

4 years, 5 months

[MM3-users] Re: Mailman3 not working following distribution upgrade from Debian 10 to 11.

by Mark Sapiro

On 2/4/22 11:15, gordon(a)dickens.com wrote: > > 173.165.166.34 - root [04/Feb/2022:12:54:06 -0500] "GET /mailman3/postorius/lists/members.georgiajets.org/ HTTP/1.1" 404 282 > 173.165.166.34 - root [04/Feb/2022:12:54:06 -0500] "GET /favicon.ico HTTP/1.1" 200 318 > > So, I get File Not Found error 404 for the web page but the favicon.ico file appears to be found successfully. Therefore, I suspect that my apache configuration file for mailman3 may be at the source of my problems and specifically the alias for the static directory. Here is my apache configuration file: It's not the alias for static > Alias /mailman3/favicon.ico /var/lib/mailman3/web/static/postorius/img/favicon.ico > Alias /mailman3/static /var/lib/mailman3/web/static > > <Directory "/var/lib/mailman3/web/static"> > Require all granted > </Directory> > > <IfModule mod_proxy_uwsgi.c> > ProxyPass /mailman3/favicon.ico ! > ProxyPass /mailman3/static ! > # ProxyPass /mailman3 unix:/run/mailman3-web/uwsgi.sock|uwsgi://localhost > </IfModule> It's the fact that ProxyPass /mailman3 unix:/run/mailman3-web/uwsgi.sock|uwsgi://localhost is commented out. This should not be commented assuming you are using uWSGI and it is listening on the socket /run/mailman3-web/uwsgi.sock -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

3 years, 5 months

[MM3-users] Re: Archive import fails: "IndexError: list index out of range"

by Mark Sapiro

On 1/4/24 9:42 PM, Stephen J. Turnbull wrote: > Dan Caballero writes: > > > I've got a message that keeps getting re-queued to the > > /hyperkitty/spool/ folder. > [...] > > "During handling of the above exception (expected atom at a start > > of dot-atom-text but found > > '[1db292f04b54496ab40544b0989a2122-JFBVALKQOJXWILKNK4YVA7CPNZSVMZLUPRCW2YLJNRKGK43UPRCXQ32TNV2HA===(a)microsoft.com]>')" > > This is arguably a HyperKitty bug. The email package is correctly > refusing to parse that as a Message-ID (it's been illegal syntax since > RFC 822 if I remember correctly), but something, presumably > HyperKitty, should catch the error and do something to sanitize it to > message-id syntax. On the other hand, I never heard of a message > trigger that exception before a couple months ago, so maybe it's the > responsibility of spam filters. This is <https://gitlab.com/mailman/mailman/-/issues/1065> fixed in Mailman 3.3.9 by <https://gitlab.com/mailman/mailman/-/merge_requests/1099>. > In a later post you say that message came back, but I suspect it isn't > the same message, but a new one likely from the same source. I agree. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

1 year, 6 months

[MM3-users] Re: Rejection based on total message size including attachments

by Mark Sapiro

On 09/19/2018 03:34 PM, leith(a)khimaira.com wrote: > As a web hosting provider I have searched and searched for a way to limit the total size of messages posted to all lists. We have moderators who just can't get it through their heads to limit the size of attachments. One list today distributed a 6 mb pdf to over 700 recipients. I have found information on creating a custom handler and some old and brief logic on limiting max_message_size from Mr. Sapiro. > > Would anyone who has developed a custom handler to reject messages based on total size, mind sharing the full file? If so, could you also state where the handler was positioned in the chain of handlers. Is this a Mailman 3 question or a Mailman 2.1 question? If Mailman 3, the architecture is quite different. This would be done with a rule, not a handler, but based on details in your question, I think you are asking about Mailman 2.1. In that case, the appropriate list is really mailman-users(a)python.org <https://mail.python.org/mailman/listinfo/mailman-users>, but since you're here, the basic info on custom handlers for MM 2.1 is at <https://wiki.list.org/x/4030615>. There is an example called MyHandler.py linked from that page. It won't do what you want but can be used as a starting point. Also, look at Mailman/Handlers/Hold.py, in particular the code following the "# Is the message too big?" for how to determine the message size. That handler will compare the size to the list's max_message_size and hold the post if it's too big. You probably want to compare the size to a fixed value or a size defined in mm_cfg.py and either reject or discard the message. If you put your handler between Moderate and Hold, it's probably safe to reject the message because it is from a member or an address in accept_these_nonmembers. Rejecting post from arbitrary non-members is a bad idea because of spam backscatter. If you need more help, please join mailman-users(a)python.org and post to that list. In any case, this could be generally useful enough that I will probably add it to the stuff at <https://www.msapiro.net/scripts/>. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

6 years, 10 months

[MM3-users] Re: Newbie Wants To Know About Multiple Lists

by Mark Sapiro

On 12/16/18 11:20 PM, Christopher Garwood wrote: > Howdy, > I am new to Mailman and there are so many things I would like to know. Before I try to find out how to install and configure the software, I need to know if, how and what is required for us to set up this software to deal with more than one email list. There's lots of documentation. Start here <http://docs.list.org/en/latest/>. > I am working with a plan to setup a CentOS VPS on which Mailman will be hosted. The objective is to host and administrate mailing list for the 3 entities that I am directly involved with. These lists must function completely unrelated to each other including the email addresses from which the emails & news letters are sent from. If this is possible can I be given information or a link to how it can be done and what is required? A single Mailman installation can support multiple lists, each with different membership, owners and moderators. For example, see <https://mail.python.org/mailman3/>. The lists can all be in the same domain or in different domains if the server supports multiple domains. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

6 years, 7 months

[MM3-users] Re: Using MySQL via unix socket

by Mark Sapiro

On 2/2/19 5:31 AM, Torge Riedel wrote: > > A call to > > mysql -umailman -p > > successfully uses the socket on the machine and logs in. Yet you haven't specified -S or --socket > I tried > > mysql+pymysql://mailman:mailmanpass@127.0.0.1/mailman3?charset=utf8&use_unicode=1&unix_socket=/var/run/mysqld/mysqld.sock > > > as the connection string but it says that mysql server is not available. > Which is the same if remove the "unix_socket" parameter. I have found > this parameter when searching the net, but I'm not sure if it is supported. I think it should be, but the actual connection is via the the sqlalchemy create_engine() function. It appears this should be passed to the pymysql Connection() function per <https://docs.sqlalchemy.org/en/rel_1_2/core/engines.html#custom-dbapi-args> What happens if in a Python 3 session, you do from pymysql import connect conn = connect('127.0.0.1', user='mailman', password='mailmanpass', database='mailman3', charset='utf8', use_unicode=1, unix_socket='/var/run/mysqld/mysqld.sock') As I understand it, that's what sqlalchemy should be doing with your database url. In my testing, the above works with or without the unix_socket= argument, so I suspect something else is wrong in your case. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

6 years, 5 months

[MM3-users] Re: FILTER_VHOST fix - was: Postorius & multi domain

by Mark Sapiro

On 4/23/20 2:48 AM, Gelpi Andrea wrote: > > I saw the change in the file. > > As I'm not a programmer How can I test your fix? > > My system was installed via Debian 10 stable repository. > > I upgrade with testing repository postorius, django and mailmanclinet > > Postorius V 1.3.2 > Django V. 2.2.12 > Django filters V 2.1.0 > Djanog mailman V 1.3.2 > mailmanclient V 3.3.0 I don't know where the Debian package puts things, but you have to find the location of postorius/views/list.py. Something like locate views/list.py should find it. Then you need to download <https://gitlab.com/mailman/postorius/-/merge_requests/500.diff> and edit it to remove the first 15 lines so the first remaining line is the diff --git a/src/postorius/views/list.py b/src/postorius/views/list.py line. Then make a backup copy of postorius/views/list.py and run the command patch postorius/views/list.py path/to/500.diff where postorius/views/list.py is actually the full path to postorius/views/list.py and 500.diff is the edited file you downloaded. Then restart uWSGI. If anything goes wrong, restore your backup copy of postorius/views/list.py and restart uWSGI. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

5 years, 3 months

[MM3-users] Re: How to strip DKIM headers on a domain level or a list level

by Mark Sapiro

On 11/23/20 8:00 PM, Joel Lord wrote: > > I've done the same thing, but Yahoo is pretty good at sanitizing their > report to make it difficult to impossible to tell who actually needs to > get ejected. I added "Sent to %(user_name)s at %(user_address)s" to the > footer and while they'd generally get the user_address wiped out, the > user_name would generally survive. For this reason, Mailman 2.1 has this feature. > # If the following is set to a non-empty string, that string is the name of a > # header that will be added to personalized and VERPed deliveries with value > # equal to the base64 encoding of the recipient's email address. This is > # intended to enable identification of the recipient otherwise redacted from > # "spam report" feedback loop messages. For example, if > # RCPT_BASE64_HEADER_NAME = 'X-Mailman-R-Data' > # a header like > # X-Mailman-R-Data: dXNlckBleGFtcGxlLmNvbQo= > # will be added to messages sent to user@@example.com. > RCPT_BASE64_HEADER_NAME = '' This hasn't been implemented for Mailman 3 but could be if there's demand. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

4 years, 8 months

[MM3-users] Re: Failed first login

by Mark Sapiro

On 8/31/20 6:59 AM, Massimo Zappalà wrote: > Hi everyone, > i installed mailman3-full on debian 10. > During installation, it asks me to enter the admin user credentials. > After configuration, I restart mailman3 and mailman3-web. > When I try to log in for the first time by entering the admin credentials, a message warns me that the credentials are incorrect. > I looked at the mailman logs, but I find nothing. > It seems that it is a django authentication problem, because even in the django admin panel I can't log in. > Looking at the mailman3web db, in the auth_user table, I see the admin user. > How can I see why it does not accept the credentials? > Thank you in advance I have no idea what the Debian package is doing, but you can see the superuser(s) with the database query select * from auth_user where is_superuser; and you can change the user's password with the django admin changepassword command and/or create a superuser with the django admin createsuperuser command. Give the commands with a --help option for more info. Since you can see the user in the auth_user table, try the django admin changepassword command to change the password. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

4 years, 10 months

[MM3-users] Re: Still unable to get mailman3 working on Ubuntu 20.04 LTS

by Mark Sapiro

On 1/6/22 6:28 PM, Stephen J. Turnbull wrote: > Michael Heseltine writes: > > > > Do you have all of these apache modules enabled, proxy.conf, > > > proxy_http.load and proxy.load, in particular proxy_http.load? > > > > Well, mostly: > > > > > # a2enmod proxy_http.load > > > Considering dependency proxy for proxy_http: > > > Module proxy already enabled > > > Module proxy_http already enabled > > > # a2enmod proxy.load > > > Module proxy already enabled > > > # a2enmod proxy.conf > > > ERROR: Module proxy.conf does not exist! > > In my Debian configuration that module is needed because it contains > the necessary directives to enable proxying. Actually though, I think that's enabled as part of proxy. > The bottom line, though, is that that error message has to do with the > internal configuration of Apache, and as far as I can see it has > nothing to do with Mailman itself. However, see https://lists.mailman3.org/archives/list/mailman-users@mailman3.org/message… I now suspect the real issue is there's no wsgi server listening on port 8000. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

3 years, 6 months

[MM3-users] Re: Help understanding why some users don't receive emails

by Mark Sapiro

On 9/21/22 14:06, Stephen Daniel wrote: > > Looking at one user in particular, I see in /var/log/syslog that postfix is > sending the email. There is nothing in bounce.log. And does the postfix log indicate the mail is accepted by the users MX? I.e. a log message with `status=sent`. > They are the only user who uses that particular email domain (embarqmail.com > ). > > Is there anything I can do to debug why they do not get the emails? Or is > this up to them and their email provider? Assuming the message is accepted by the user's MX, the user can check a spam/bulk/junk folder and if the messages aren't there, complain to their ISP about messages being silently discarded. Your log entries showing the message successfully delivered to the MX can provide evidence. Other than that, you need to ensure that your outgoing messages are DKIM signed, and that your outgoing MTA server has full circle DNS <https://en.wikipedia.org/wiki/Forward-confirmed_reverse_DNS>. It can also possibly help for the server's domain to publish a DMARC policy even if that is `p=none` -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

2 years, 10 months

[MM3-users] Re: List users from different domains getting blocked.

by Mark Sapiro

On 12/14/23 14:19, bob B via Mailman-users wrote: > I am running MM3 and have a list that has users from different organizations and email domains. > > So let's say I am orga.com and hosting mailman > Bill is borg.com > Cindy is mew.com > > When Bill sends a message to my MM3 list it gets sent out, However, cindy's email system rejects it cause it sees the message is sent from Bill(a)borg.com but it came from my MM3 server which is under orga.com, so it rejects it. You need to apply DMARC mitigations. Normally mitigations are only applied if the From: domain publishes DMARC p=reject or p=quarantine, but recently this has become problematic. For example, gmail.com pubpishes DMARC p=none yet google servers effectively apply p=reject to mail From: gmail.com. Mailman >= 3.3.9 and Postorius >= 1.3.9 allow setting dmarc_addresses to a list of email addresses and regexps matching addresses and domains to which DMARC mitigatations will be applied regardless of puublished policy. For older releases and for cases where domains are not known in advance one can always set dmarc_mitigate_unconditionally True to apply mitigations to all mail. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

1 year, 7 months

[MM3-users] Re: Clearing webserver cache - Nginx

by Mark Sapiro

On 8/5/24 07:04, Odhiambo Washington via Mailman-users wrote: > I am having an ugly-looking https://mm3-lists.kictanet.or.ke/archives/ page > because of cached content: > > ``` > 2024/08/05 17:00:31 [error] 694#694: *9562 open() > "/opt/mailman/mm/static/CACHE/css/output.6dab123e4897.css" failed (2: No > such file or directory), client: 172.70.111.39, server: > mm3-lists.kictanet.or.ke, request: "GET > /static/CACHE/css/output.6dab123e4897.css HTTP/2.0", host: " > mm3-lists.kictanet.or.ke", referrer: " > https://mm3-lists.kictanet.or.ke/archives/" > 2024/08/05 17:00:31 [error] 694#694: *9564 open() ... > That IP address appearing there is not my actual IP, but Cloudflare's, > which is acting as proxy for the site. I **think** you may have something like location /static/ { alias /opt/mailman/web/static/; } in nginx on cloudflare, but since cloudflare is not the Mailman server, /opt/mailman/web/static/ doesn't exist there. I can't be sure on how to configure this without knowing the details of how requests get from cloudflare to mm3-lists.kictanet.or.ke and from there to the wsgi server. However, if cloudflare just proxies the request for /static/ to the Mailman server and nginx on that server has the above `location` directive, that may do it. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

11 months, 3 weeks

[MM3-users] Re: Update a member's display_name

by Mark Sapiro

On 7/8/20 11:08 AM, rongxinliu(a)cs50.harvard.edu wrote: > Hi Mark, I'm not the only developer on this list. > I would like to know if there is a way to update the display_name of the member in a mailing list via REST API? > > Having tried updating the display_name of a user but did not seem to work. The display_name did change, but the mailing list still displaying the old name. By "the mailing list still displaying the old name" do you mean the membership list in Postorius or something else? And to what REST endpoint did you post the display_name? The bottom line is a user can have multiple display names. Each user has a User record with a display_name attribute, one or more 'addresses', one of which is it's preferred_address. Each one of the addresses is an Address record with it's own email address and display_name. Now, a member of a list can be either a User or an Address. If it's a User, its display_name is that of the user and it's email address is that of the preferred_address, but if it's an Address, its email and display_name are those of the Address. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

5 years

[MM3-users] Re: Hyperkitty no longer archiving

by Mark Sapiro

On 11/22/22 13:21, Hirayama, Pat wrote: > > This is the end of settings.py: > # > # HyperKitty-specific > # > > # Only display mailing-lists from the same virtual host as the webserver > FILTER_VHOST = False > > > POSTORIUS_TEMPLATE_BASE_URL = 'http://localhost:8000' > > > try: > from settings_local import * > except ImportError: > pass > > I'm assuming that is close enough to the end? Yes. > So, looking at the Django container, I see this process popping up periodically. I think that means that it is running from the directory specified by "--pythonpath"" (which is where settings.py and settings_local.py are located). I am further assuming that "--settings settings" is ok without a .py extension. > > 53972 mail 0:00 {django-admin} /usr/local/bin/python /usr/local/bin/django-admin runjobs quarter_hourly --pythonpath /opt/mailman/mailman-suite/mailman-suite_project/. --settings settings > 53978 mail 0:00 {django-admin} /usr/local/bin/python /usr/local/bin/django-admin runjobs minutely --pythonpath /opt/mailman/mailman-suite/mailman-suite_project/. --settings settings Yes, it is `--settings settings` without a .py extension because that's the name that Python imports. However, those are from cron jobs and say nothing about the wsgi process that runs the web UI and the hyperkitty api. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

2 years, 8 months

[MM3-users] Re: Mailman sends Welcome emails, but does not send [List] postings emails

by Mark Sapiro

On 6/7/23 21:30, Nelson Strother wrote: > > Interestingly if I use `systemctl start mailman` thus far the results are: > - the wall clock duration is shorter than when `mailman start` is issued > from the command line, and > -- either that all of the runners remain present once they are started, > -- or I can see in syslog a traceback from each missing runner process > starting from .../mailman/.local/bin/runner and ending with > "flufl.lock._lockfile.NotLockedError: Already unlocked". > > I do not yet understand how to make use of these clues, but at least one > can see an epitaph from each deceased process. The NotLockedError is most likely because a lock has been set with a lifetime and its lifetime has passed and some other process has broken the lock before the original process tries to unlock it. This probably has something to do with the long times to do start/restart. However, the only lock that should be involved is the master lock set by the master watcher, so I don't really understand what might be happening. What is the full traceback from these errors? -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

2 years, 1 month

[MM3-users] Re: How to Personalise Email Content with Usernames in Mailman 3

by Mark Sapiro

On 11/17/23 07:03, Zaheer Abbas wrote: > Hello, > > I am currently working with Mailman 3 and looking to enhance the > personalization of the emails sent through our mailing list. Specifically, > I wish to include the recipient's username in the email content, similar to > a personalized greeting (e.g., "Dear [username]"). > > My questions are: > > 1. Does Mailman 3 support the inclusion of personalized data (like > usernames)? You can enable full personalization which will put the recipient's address in the To: header. In addition both Individual and Full personalization allow personalized information in message headers and footers, see https://docs.mailman3.org/projects/mailman/en/latest/src/mailman/rest/docs/…. This could be used to add a personalized header, but depending on the MIME structure of the message after content filtering, these may be added as separate MIME parts rather than prepended/appended to the message body. See https://wiki.list.org/x/4030707 for more on this. Mailman has no facility other than the above for modifying message bodies. > 2. If yes, how can I do that? Is there any placeholder for that? See the above links. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

1 year, 8 months

[MM3-users] Re: HSTS Missing from HTTPS Server

by Mark Sapiro

On 11/4/24 12:06, jkoral(a)luc.edu wrote: > > My question is, how can enable the Virtual Host for 443 on the server? Or can that configuration be located somewhere else? Or can I just create a new Virtual Host with the added configuration for mailman? For example, this is /etc/apache2/sites-enabled/020-mailman.conf on lists.mailman3.org ``` <VirtualHost *:80> RewriteEngine on RewriteCond %{HTTPS} off [NC] RewriteRule ^/(.*) https://%{HTTP_HOST}/$1 [L,R=permanent] RewriteCond %{SERVER_NAME} =lists.mailman3.org RewriteRule ^ https://%{SERVER_NAME}%{REQUEST_URI} [END,NE,R=permanent] </VirtualHost> <VirtualHost *:443> SSLEngine on ServerName lists.mailman3.org Alias /static/ /opt/mailman/mm/static/ <Directory "/opt/mailman/mm/static/"> Require all granted </Directory> <IfModule mod_headers.c> RequestHeader unset X-Forwarded-Proto <If "%{HTTPS} =~ /on/"> RequestHeader set X-Forwarded-Proto "https" </If> </IfModule> <IfModule mod_proxy.c> ProxyPreserveHost On ProxyPass "/postorius" "http://127.0.0.1:8000/postorius" ProxyPass "/hyperkitty" "http://127.0.0.1:8000/hyperkitty" ProxyPass "/accounts" "http://127.0.0.1:8000/accounts" ProxyPass "/admin" "http://127.0.0.1:8000/admin" ProxyPass "/user-profile" "http://127.0.0.1:8000/user-profile" ProxyPass "/mailman3" "http://127.0.0.1:8000/mailman3" ProxyPass "/archives" "http://127.0.0.1:8000/archives" ProxyPassMatch "^/$" "http://127.0.0.1:8000/mailman3" </IfModule> Include /etc/letsencrypt/options-ssl-apache.conf SSLCertificateFile /etc/letsencrypt/live/www.list.org/fullchain.pem SSLCertificateKeyFile /etc/letsencrypt/live/www.list.org/privkey.pem </VirtualHost> ``` -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

8 months, 4 weeks

[MM3-users] Re: getting all registered users and their lists

by Mark Sapiro

On 1/27/22 00:06, IOhannes m zmoelnig wrote: > On 1/26/22 17:19, Mark Sapiro wrote: >> On 1/26/22 00:16, IOhannes m zmoelnig wrote: > >>> is there a scriptable way to get all users that have registered in my >>> mailman3 (3.2.1) instance and which lists they are subscribed to (if >>> any)? >> >> It's in Postorius >= 1.3.6 > > i'm not entirely sure how this relates to the above (working) snippet. > > is it, that with postorious>=1.3.6 i would be able to perform that task > from the webinterface, rather than having to resort to 'mailman shell'? Not really. In Postorius 1.3.6 there is a `Users` view that will display users and their create date and it is ordered by create date. Then each user's entry has a `Manage` button which goes to a page which among other things, lists their subsubscriptions. So, you could find the users of interest via this interface, but unless you developed a screen scraping script, it would be a multi-step manual process. > in any case, this might be related: > before i received your email with the 'mailman shell' snippet, i scraped > some bits of my django knowledge and visited > <https://lists.example.com/admin> (the django admin interface), and > noticed the list of users there. > so i scraped that page for users (assuming that this was the list of > registered users). > then i get myself a list of all subscribers in my instance, using a > crude `mailman lists | grep "@" | while read l; do mailman members > "${l}"; done | sort -u`. > finally i calculated the difference between the two sets. > > interestingly the results are quite different from the ones i got from > your script: both sets contain users that are not in the other set. > how come? My script deals with Mailman users who may or may not be Django users. They won't be if they never signed up on the web UI. On the other hand, anyone can go to the web UI and sign up and become a Django user even if they never join any list and become a Mailman user. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

3 years, 6 months

[MM3-users] Re: Cron of Mailman notify is not working

by Mark Sapiro

On 1/29/20 12:01 AM, Shashikanth Komandoor wrote: > > I observed that the "notify" subcommand can be run only while I am in the /var/lib/mailman/mailman directory where my mailman core is installed (through GITHUB). I think you mean GitLab, but in any case, the reason for this is you have to be in that directory for the mailman command to find your mailman.cfg file. I deal with this by creating a separate bin/ directory for these commends, in my case /opt/mailman/mm/bin because Mailman is installed in /opt/mailman/mm/, and my /opt/mailman/mm/bin/mailman file contains #!/bin/bash exec /opt/mailman/mm/venv/bin/mailman \ -C "/opt/mailman/mm/mailman.cfg" \ $@ > And moreover I installed all the components of the mailman using the root user only and so there is no mailman user on my machine. > > And hence I wrote the below entry in the crontab directly and found it is working fine. > > 00 08 * * * cd /var/lib/mailman/mailman; /var/lib/mailman/venv3/bin/python /var/lib/mailman/venv3/bin/mailman notify The /var/lib/mailman/venv3/bin/python is probably not needed as /var/lib/mailman/venv3/bin/mailman should contain the shebang line #!/var/lib/mailman/venv3/bin/python -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

5 years, 6 months

[MM3-users] Re: 12/24 hour clock ?

by Mark Sapiro

On 11/28/20 7:18 AM, Per Jessen wrote: > Hi all > > at openSUSE we have recently migrated to mailman3, and we seem to be > having a couple of teething issues: > > a) hyperkitty and postorius both seem to insist on using a 12 hour clock > format, no matter what the user has set. Is there a way of changing it > system-wide ? This is a Django setting. Try TIME_FORMAT = 'H' in settings_local.py. See https://docs.djangoproject.com/en/3.1/ref/settings/#time-format and https://docs.djangoproject.com/en/3.1/ref/templates/builtins/#std:templatef… > b) overriding the default set of footers - despite having set the > template "list:member:regular:footer" to empty for a specific list, I > am still getting the default footers attached. How did you set the template to empty? it works for me in Postorius from the HEAD of the gitlab branch and should work with any Postorius >= 1.3.0 Note that when you set a template in Postorius, it sets the URI in Mailman core for that template to https://<HOST>/<PATH>/api/templates/list/<LIST-ID>/<TEMPLATE-NAME> where the https://<HOST>/<PATH> part is the Django setting for POSTORIUS_TEMPLATE_BASE_URL. Is that URI accessible from the Mailman core host machine? -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

4 years, 8 months

[MM3-users] Re: Mailman3 doesn't send Mails to users (Debian12 with Plesk)

by Mark Sapiro

On 6/6/24 11:22 AM, Marcus Bautze wrote: > Hi, > > Sorry, that was a wrong information in my mail . I already changed the base url to your suggestion as I wrote the mail (and forgot to change it in the mail) but i‘m getting the same error :-( I have some questions. At https://lists.mailman3.org/archives/list/mailman-users@mailman3.org/message… you say "Thank you, archiver is now starting. Now there's another problem:". Does this mean messages are being archived in HyperKitty or are they accumulating in (I think the Debian path is) /var/lib/mailman/archives/hyperkitty/spool/? At https://lists.mailman3.org/archives/list/mailman-users@mailman3.org/message… you say "The Keys are in mailman-hyperkitty.cfg and mailman-web.py and this are the same Keys." To be clear, in mailman-web.py the key is quoted and in mailman-hyperkitty.cfg it is not. E.g., in mailman-hyperkitty.cfg api_key: Some_key and in mailman-web.py MAILMAN_ARCHIVER_KEY = 'Some_key' What are your versions for mailman-hyperkitty and HyperKitty? HyperKitty >= 1.3.5 is incompatible with mailman-hyperkitty < 1.2.0 and if I recall corectly, at least one Debian package was shipped with this incompatibility -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

1 year, 1 month

[MM3-users] Re: how to mass change users options ?

by Mark Sapiro

On 4/30/22 02:13, Yvan ♞ wrote: > Hello, I transfered a mailing list from mailman to mailman 3, how can I > change all users options in mass ? > delivery status and receive own posting are blank by default (250 > members is too long one by one..) The `blank` value for for a setting only means that the user has no list based setting for that property. This means either the user's address based value if any applies, otherwise the Mailman global value applies. That said, if you imported the list with `mailman import21`, the Mailman 2.1 settings should have been imported. Finally, you could change a setting for all list members using `mailman shell`. For example ``` $ bin/mailman shell -l list(a)example.com Welcome to the GNU Mailman shell Use commit() to commit changes. Use abort() to discard changes since the last commit. Exit with ctrl+D does an implicit commit() but exit() does not. The variable 'm' is the list(a)example.com mailing list >>> for member in m.members.members: ... member.preferences.delivery_status = DeliveryStatus.enabled ... member.preferences.receive_own_postings = True ... >>> commit() ``` But this is probably unnecessary as the global settings probably apply and are probably set appropriately. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

3 years, 3 months

[MM3-users] Re: New messages are not archived - may be a path issue

by Mark Sapiro

On 10/20/23 05:31, Wikinaut wrote: > > The built-in template (mailman-hyperkitty.cfg) perhaps needs to be fixed as well (hyperkitty wrong, archive correct)? > # However, if your Mailman installation is accessed via HTTPS, the URL needs > # to match your SSL certificate (e.g. https://lists.example.com/hyperkitty) > # base_url: http://localhost/mailman3/hyperkitty/ > > this line perhaps needs to be changed to > # base_url: http://localhost/mailman3/archives/ I agree that this is confusing because some of our docs recommend that urls.py contain ``` path('postorius/', include('postorius.urls')), path('hyperkitty/', include('hyperkitty.urls')), ``` and docs recommend ``` path('mailman3/', include('postorius.urls')), path('archives/', include('hyperkitty.urls')), ``` I.e. following the first suggestion, you would access Postorius and HyperKitty via `https://www.example.com/postorius/` and `https://www.example.com/hyperkitty/` respectively, but with the second, the urls would be `https://www.example.com/mailman3/` and `https://www.example.com/archives/` respectively. I'm going to address this by adding ``` path('postorius/', include('postorius.urls')), path('hyperkitty/', include('hyperkitty.urls')), ``` to mailman_web's urls.py so the either urls will work. See https://gitlab.com/mailman/mailman-web/-/issues/24 -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

1 year, 9 months

[MM3-users] Re: Welcome message template errors

by Mark Sapiro

On 4/23/24 23:20, Mark wrote: > > The issue is: > > A. With POSTORIUS_TEMPLATE_BASE_URL = > 'http://<REDACTED-PUBLIC-IPv4-ADDRESS>', it sends an empty Welcome > message when a new member is added. > > B. With POSTORIUS_TEMPLATE_BASE_URL = 'https://MYDOMAIN.COM', no Welcome > message gets sent and a timeout error appears on the webpage, although > the new member gets added. I think the difference is http vs https, not the IP vs domain name. Also, with either of these, you are addressing the web server and not the wsgi server. The wsgi server is presumable listening on port 8000 and in the beginning when you had the default POSTORIUS_TEMPLATE_BASE_URL = 'http://localhost:8000' you also got the timeout error and you also got an email to the site owner with 'Site matching query does not exist.' for 'Request URL: http://<REDACTED-PUBLIC-IPv4-ADDRESS>/mailman3/lists/' This leads me to think 'localhost' resolves to <REDACTED-PUBLIC-IPv4-ADDRESS> instead of 127.0.0.1. Also, <REDACTED-PUBLIC-IPv4-ADDRESS> is not in ALLOWED_HOSTS. Have you set ALLOWED_HOSTS in your settings.py? It should be something like ``` ALLOWED_HOSTS = [ "localhost", # Archiving API from Mailman, keep it. "127.0.0.1", "<REDACTED-PUBLIC-IPv4-ADDRESS>", "MYDOMAIN.COM", ] ``` -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

1 year, 3 months

[MM3-users] Re: Re-enable member after bouce processed disable

by Mark Sapiro

On 1/5/22 12:48 AM, Henrik Rasmussen wrote: > > It seems that I was right, however, I need to re-enable all members on a single list who have been disabled due to bounces. > > Using the `mailman members` command it is possible to list such members, however how do I enable them using either the `mailman shell` command (available) og the mailmanclient (needs first to be installed in the docker container)? There is a `mailman shell` session in the post at https://lists.mailman3.org/archives/list/mailman-users@mailman3.org/message… that will re-enable all bounce disabled members on all lists. To do just a single list do something like ``` $ bin/mailman shell -l list.example.com Welcome to the GNU Mailman shell Use commit() to commit changes. Use abort() to discard changes since the last commit. Exit with ctrl+D does an implicit commit() but exit() does not. The variable 'm' is the list.example.com mailing list >>> for member in m.members.members: ... if member.delivery_status == DeliveryStatus.by_bounces: ... member.preferences.delivery_status = DeliveryStatus.enabled ... print("Re-enabled delivery for bouncing member {} on list {}" ... .format(member.address.email, mlist.list_name)) ... <print output here> >>> commit() >>> exit() ``` -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

3 years, 6 months

[MM3-users] Re: (venv) mailman-web migrate (as explained) fails with "django.core.exceptions.ImproperlyConfigured: allauth.account.middleware.AccountMiddleware must be added to settings.MIDDLEWARE"

by Mark Sapiro

On 10/20/23 05:43, Wikinaut wrote: > it has: > > # Mailman Web configuration file. > # /etc/mailman3/settings.py > > # Get the default settings. > from mailman_web.settings.base import * > from mailman_web.settings.mailman import * As I said, I don't know what caused your original authentication failure or what fixed it, but it wasn't the addition of ``` MAILMAN_REST_API_USER = 'restadmin' MAILMAN_REST_API_PASS = 'restpass' ``` to /etc/mailman3/settings.py. You can verify this by activating your venv and doing ``` >>> MAILMAN_REST_API_USER Traceback (most recent call last): File "<stdin>", line 1, in <module> NameError: name 'MAILMAN_REST_API_USER' is not defined >>> MAILMAN_REST_API_USER Traceback (most recent call last): File "<stdin>", line 1, in <module> NameError: name 'MAILMAN_REST_API_USER' is not defined >>> MAILMAN_REST_API_PASS Traceback (most recent call last): File "<stdin>", line 1, in <module> NameError: name 'MAILMAN_REST_API_PASS' is not defined >>> from mailman_web.settings.mailman import * >>> MAILMAN_REST_API_USER 'restadmin' >>> MAILMAN_REST_API_PASS 'restpass' >>> ``` -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

1 year, 9 months

[MM3-users] Re: add custom styles to postorius list creation menu

by Mark Sapiro

On 5/2/22 06:39, Michael Hazan wrote: > Hello, > > When creating a list via postorius, i can see 3 style appear (announce, discussion, private). Is it possible to either create a new style and add it to that list or modify one of those style ? See this thread https://lists.mailman3.org/archives/list/mailman-users@mailman3.org/thread/… > Also, coming from mailman2, you could directly set the members post moderation action when subscribing, now you can only do it via the user menu, is there a way to automate this in mm3 ? I don't understand what you mean here. In MM 2.1 a member had only one setting - moderated yes or no and if moderated there was a list setting for the moderation action. There was also a list setting for whether new members were moderated by default. How were you setting the new member's moderation when subscribing? You are correct that when subscribing a new member, the member gets the default action for members and different action has to be set in a separate operation. I don't think this is significantly different from MM 2.1. Also, can you describe what you mean by "automate this". -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

3 years, 2 months

[MM3-users] Re: Query on how to change moderator message to non-members

by Mark Sapiro

On 7/1/18 10:55 PM, robert.kydd(a)gmail.com wrote: > I have had a look in cPanel File Manager but did not find anything. Is it possible that these files are stored at the server level which would then be under control of the ISP hosting the website ? What Mailman is this? If this is in fact Mailman 3, the var/ directory referred to at <https://gitlab.com/mailman/mailman/issues/426#note_43861203> is probably not /var/, It is Mailman's var directory as configured by var_dir in mailman.cfg and the template you're interested in is list:user:notice:hold.txt. If this is cPanel's Mailman 2.1, you can edit the template through Mailman's web admin interface. Go to "Edit the public HTML pages and text files" in the upper right and then select "User notice of held post". However, in either case, the "post by nonmember" reason is a replacement, so if you edit the template to put different text there, it will apply to all reasons. One way around this is to make an English language i18n message catalog which translates the "Post by non-member to a members-only list" message to something else. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

7 years

[MM3-users] Re: Archiver Hyperkitty Not Found

by Mark Sapiro

On 3/3/23 15:16, pat(a)suwalski.net wrote: > Hello, > > I upgraded from Debian 10 to 11 today, and hyperkitty archiver stopped working. > >>From the mailman.log: > > AssertionError: Archiver not found: hyperkitty > > I followed along on a related thread: https://lists.mailman3.org/archives/list/mailman-users@mailman3.org/thread/… > > I don't see any resolution to that thread. The only thing that worked for me was removing the hyperkitty rows from the listarchiver table in the database. But of course, now I don't have working archiving. There is a resolution of sorts there: It was setting ``` [archiver.hyperkitty] enable: no ``` in mailman.cfg, restarting Mailman core and then setting it back to `enable: yes`. > Any ideas? How would I make the archiver available? This django stuff is infuriatingly difficult compared to other frameworks. Do you have ``` [archiver.hyperkitty] class: mailman_hyperkitty.Archiver enable: yes configuration: /opt/mailman/mm/mailman-hyperkitty.cfg ``` in your mailman.cfg where `/opt/mailman/mm/mailman-hyperkitty.cfg` is the actual path to mailman-hyperkitty.cfg and do you have in that file ``` [general] base_url: http://example.com/hyperkitty/ api_key: key matching the MAILMAN_ARCHIVER_KEY set in Django's settings. ``` Actually, even if these latter settings were wrong, the error would be different. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

2 years, 4 months

[MM3-users] Re: Apache shows No Lists

by Mark Sapiro

On 5/22/23 12:57, ceo(a)sartrack.nz wrote: > Postorius is running. > I just removed all mailing lists, and created new empty ones. > These now show in the Apache2 website, via the link to Postorius I assume. > Then I imported members from the old mailing list "Templist" into the new one. > Now the "templist" is no longer visible on the website! Did you do this import with `mailman import21`? If so, did you run it as the Mailman user and not as root? > It is here: > https://sartrack.nz/mailman3/ > Note that the existing list "Templist" is missing. https://sartrack.nz/mailman3/ is missing CSS. Apache doesn't know how to access /static/ or you never built it with the django admin collectstatic command. > Now I need to login as 'superuser' or whatever it is called, so I can manage all the lists. > Where do I set up my login and password for access via the web browser? You create a superuser with the django admin createsuperuser command. How you invoke the django admin command depends on how you installed Mailman. Please tell us how you installed Mailman so we don't have to guess about this and other things. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

2 years, 2 months

[MM3-users] Re: bounce handling in mm3 available?

by Mark Sapiro

On 1/24/22 05:20, Stefan Bauer wrote: > Hi, > > what happens with bounce-mails that are processed by mailmans postfix > instance? What Mailman core version? Prior to 3.3.1, bounce events were recorded in the bounceevent table but not otherwise processed. Subsequent to that, they are scored if recognized and what notices are sent depends on list settings. > Jan 24 10:46:03 l01 postfix/lmtp[8432]: DF4AE600623: > to=<company-newsletter-bounces(a)lists.company.tld>, > relay=127.0.0.1[127.0.0.1]:8024, delay=0.03, delays=0.01/0.01/0/0.01, > dsn=2.0.0, status=sent (250 Ok) > > But the list-owner does not get any notification, nor do i see any > bounce-information in the pistorious-gui. What are you expecting to see in Postorius? All you can see is if a member's delivery status is Disabled by Bounces. Things like Bounce score are not available in Postorius. What notices are sent depends on the settings in the Bounce Processing section in Postorius. For example, if your message logged by postfix was spam to the <company-newsletter-bounces(a)lists.company.tld> address that is not recognized as a DSN and `Forward unrecognized bounces` is Discard, the message will be silently discarded. Also, if that message was a non-fatal, message-delayed type of DSN, it will be ignored. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

3 years, 6 months

[MM3-users] Re: If I manually add email addresses to a list and then delete them, how do i permanent delete them from Users?

by Mark Sapiro

On 4/25/23 3:02 AM, Andrew Hodgson wrote: > Mark London wrote: > >> I added some email addresses manually to a mailing list and then removed them. But now I can't find how to totally remove them from mailman3. They show up in the Users list. > > You may find this script useful: > https://www.msapiro.net/scripts/clean_users.py That script will only delete Django users and optionally the Mailman user corresponding to the deleted Django user. I think the OP wants to delete Mailman users that are not members of any lists even if there is no corresponding Django user. That script won't do that. Mailman Users can be deleted from the Postorius Users view, but it's a one at a time process, and there's no indication that the User has memberships unless you first go to `Manage` to check. I will modify the script to enable deleting Mailman users with no memberships even if they aren't Django users and post here when it's done. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

2 years, 3 months

[MM3-users] Re: List nonmembers

by Mark Sapiro

On 7/8/19 6:50 PM, Andrew Bernard wrote: > Under Users for a list, there is a category for List nonmembers. What does this term mean? [It's rather curious philosophically when you think about it, because I would have thought that the category [all humans - list members] fits that category term.] > > But seriously, I don't understand it. I have several email addresses there and I don't know what do with them, if anything. Non-members are email addresses that are not members but have some connection to a list. Mailman 2.1 lists have attributes named accept_these_nonmembers, hold_these_nonmembers, reject_these_nonmembers and discard_these_nonmembers. These are lists of non-member addresses and regexps matching addresses, and posts from a non-member address in one of the lists will be accepted, held, rejected or discarded respectively. Mailman 3 does this differently. An address can be a non-member and that non-member record has a moderation setting to determine how a post from that non-member is handled. An address can become a non-member by being imported from Mailman 2.1 or by posting to the list. If a non-member posts to the list, a non-member record is created regardless of the disposition of the post. See <https://mailman.readthedocs.io/en/latest/src/mailman/chains/docs/moderation…>. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

6 years

[MM3-users] Re: Strange password problem

by Mark Sapiro

On 9/22/23 09:50, Eggert Ehmke via Mailman-users wrote: > > -I try to log in as admin > -rejected because of wrong password (password is stored in browser) > -used forgotten password function, using my email address > -reset password to the stored value from browser > -login works now. Works also when logged out/in again > -however, the next day this behaviour repeats. So apparently the issue is that although you reset your login password to a known value, it doesn't stick and reverts overnight to some other value. > Can we try to catch this bug now, which is most likely in my configuration? Which is: > > Debian 12.1 Bookworm > Mailman in venv running with dedicated mailman user Is this the Debian package or installed following https://docs.mailman3.org/en/latest/install/virtualenv.html or something else. All I can think of is some cron may be resetting the password. If that's it, it's probably a Debian thing. Can you see if there's any cron running the Django admin changepassword command? If you set the password with the Django admin changepassword command, does it stick? Can you create another superuser with the Django admin createsuperuser command and does that user's password stick? -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

1 year, 10 months

[MM3-users] Re: Solved Re: after last upgrade I've got a CSRF failed verification

by Mark Sapiro

On 10/27/22 03:04, Guillermo Hernandez (Oldno7) via Mailman-users wrote: > > Since I made all the mailman3 enviroment wide server open (not using > venv) I already executed the migration doing a > > su -m mailman3 -c "python3 manage.py collectstatic" > su -m mailman3 -c "python3 manage.py migrate" > > Being mailman3 the user who runs all of mailman3 related stuff > > I had problems doing the migrate part, showing an error of lacking > perms. I solved it running the migration under root (I know it could be > trouble after, but...) It would be much better to identify the problem files/directories and adjust their ownership. It is possible that running stuff as root in the past left things owned by root causing the issue. > > After that the compress part showed errors... but the postorius web > maintenance was working well except for the CSRF error verification. I > will trace the compress errors and will try to find the problem later. Do you have `COMPRESS_ENABLED = True` in your settings? If that's not the issue, it may be in the settings for COMPRESS_PRECOMPILERS -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

2 years, 9 months

[MM3-users] Re: Starting from scratch

by Mark Sapiro

On 04/09/2018 10:01 AM, Benjamin Ragheb wrote: > > The two alternatives I am weighing right now are: > > 1. Stay the course, and see if I can fix this by learning more about Python, Pip, Django and tinkering with the configuration. > > 2. Ride the bleeding edge, and install postorius and hyperkitty from git. (I'll probably stick with Mailman Core from Pip, since that seemed to work just fine.) > > Right now I'm leaning toward #2, because it means I'll be in a better position to patch things if I need to, and it seems like Mailman 3 is still rough enough that I will need to. Also, it looks like the master branches of both postorius and hyperkitty are now Python 3 compatible, which could save me from managing two different versions of Python. > > Which is the least crazy option? If you are comfortable with 2), I think that's the better option. After PyCon in May when we release 3.2 and the rest of the suite, you should be able to install everything with pip and have it work. I'm not sure what the issues are now with a pip install. There is a 3rd option. See <https://packages.debian.org/stretch-backports/mailman3-full> -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

7 years, 3 months

[MM3-users] Re: Rerun emails in the queue/bad

by Mark Sapiro

On 11/20/18 8:20 AM, Kaleb White wrote: > How do I run the emails in the queue/bad again? I need to see why the failed and get them to work. Thanks. There should be error messages in the 'mailman' log for each preserved file. These would be like 'Skipping and preserving unparseable message: ....psv' 'Unpickling .bak exception: <error msg> Preserving file: ....psv' '.bak file max count, preserving file: ....psv' The first of these indicates an error in retrieving the message from it's original queue. The others have to do with how message loss is avoided. When a .pck file in a queue is processed, it is read and renamed to a .bak. Then if processing is interrupted by a crash, upon restart the .bak file will be processed. The second error above indicates a problem in unpickling the .bak and the third says the same .bak has already been tried too many times. You may be able to examine the files by pointing 'mailman qfile' at them. If that fails, either 'cat' or 'strings' may produce something useful. If they appear to be good, you can reprocess them by renaming to .pck, moving to queue/shunt and running 'mailman unshunt'. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

6 years, 8 months

[MM3-users] Re: translating templates (was: Re: spanish localization)

by Mark Sapiro

On 2/16/21 2:59 AM, Guillermo Hernandez (Oldno7) via Mailman-users wrote: > > /usr/local/lib/python3.7/site-packages/mailman # ls templates/es > domain:admin:notice:new-list.txt list:member:regular:header.txt > help.txt list:user:action:invite.txt > list:admin:action:post.txt list:user:action:subscribe.txt > list:admin:action:subscribe.txt list:user:action:unsubscribe.txt > list:admin:action:unsubscribe.txt list:user:notice:goodbye.txt > list:admin:notice:disable.txt list:user:notice:hold.txt > list:admin:notice:removal.txt list:user:notice:no-more-today.txt > list:admin:notice:subscribe.txt list:user:notice:post.txt > list:admin:notice:unrecognized.txt list:user:notice:probe.txt > list:admin:notice:unsubscribe.txt list:user:notice:refuse.txt > list:member:digest:header.txt list:user:notice:rejected.txt > list:member:digest:masthead.txt list:user:notice:warning.txt > list:member:generic:footer.txt list:user:notice:welcome.txt > > But mailman isn't aware of them (I restarted the server). May I have to > execute some process as "compilemessages" to mailman be capable to use > them? Now mailman is using the "en" templates for sure. Does Mailman have permission to search the templates/es directory and read the templates? Are there overriding templates in Mailman's var/templates/ directory? Are you sure your tests access the templates in a Spanish language context? -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

4 years, 5 months

[MM3-users] Re: Debian Package Install - Ownership and Permissions

by Mark Sapiro

On 4/7/23 14:58, Dave Hall via Mailman-users wrote: > On Fri, Apr 7, 2023 at 5:47 PM Mark Sapiro <mark(a)msapiro.net> wrote: > >> On 4/7/23 13:43, Odhiambo Washington wrote: >>> On Fri, Apr 7, 2023 at 8:09 PM Andreas Schamanek < >>> as2020+github(a)fam.tuwien.ac.at> wrote: >>>> >>>> I am running MM2 und 3 in parallel without problems, I do have a >>>> special MTA setup though (Sendmail). Can't speak for Postfix etc.. >>>> >>> >>> Your setup can handle list1(a)domain1.name via MM2 and MM3, >> simultaneously, >>> right? :-) >> >> It is possible to have an MM2 `list1` and an MM3 `list1(a)domain1.name` >> simultaneously on the same server, but what is the MTA going to do with >> incoming mail to `list1(a)domain1.name`. I suppose it could deliver it to >> both MM2 and MM3, but I don't see any practical use case for this. >> >> I run Exim4 on my servers so I can comment from this perspective. Exim > had a concept of 'routers' which are stacked and processed in order of > appearance. There is a router for MM2 and one for MM3 that pick off > incoming mail based on whether the email address matches a list name. WIth > this setup, you could have the same list name in both MM2 and MM3, but it > would go to whichever router appears first and the second one would never > see it. So I have the MM3 router first, and as I migrate my lists from > MM3 the email for those lists no longer gets to MM2. And mail.python.org which currently has 189 Mailman 3 lists and 223 Mailman 2.1 lists does something very similar in it's Postfix configuration and also prefers Mailman 3. However, maybe I misunderstood Odiambo's question about simultaneous support for a Mailman 2.1 and Mailman 3 list of the same name. We could certainly have that on mail.python.org, and even have both MM 2.1 and MM 3 web UIs for those lists, but as soon as the MM 3 list is created during migration, mail is only delivered to the MM 3 list. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

2 years, 3 months

[MM3-users] Re: Mailman 3 Social Auth Configuration

by Mark Sapiro

On 10/12/2016 08:09 PM, emonical(a)misfitlabs.org wrote: > > I am a new Mailman 3 user and have just gotten an environment configured using the Mailman bundler. The setup is almost complete, but, there's still Persona authorization. Mailman bundler installs the latest releases from PyPI, but those are behind the latest development branches. > I was reading the your setup here from (https://wiki.list.org/DOC/Mailman%203%20installation%20experience) and was wondering if someone might have some insight as to how to remove the Persona log in to keep it locked down entirely to localized user registration. With the latest gitlab branches, you will have Django Allauth. This includes login by verified email and password, and if you choose to not use any of the supported Oauth2 providers, that will be the only method. To do this without waiting for the Mailman 3.1 release and an updated bundler, you will need to install the latest master branches from <https://gitlab.com/groups/mailman> for Mailman, Postorius, HyperKitty, MailmanClient, django-mailman3 and HyperKitty - Mailman plugin. The wiki article <https://gitlab.com/groups/mailman> will give some insight into the process, but it isn't fully current. I will be updating it. Meanwhile, ask your specific questions here and we will try to help. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

8 years, 9 months

[MM3-users] Re: Mailman archive issues (Hyperkitty)

by Mark Sapiro

On 6/8/21 10:50 PM, Kimmo L wrote: > Okay, i will try to explain the overall situation as far as I can understand: > Its a CentOS 7 host without any virtualenv. Mailman_suite_project git repo has been cloned to the host and after that some packages have been installed with the help of pip: > > # pip2 install django==1.10 postorius==1.1.0 hyperkitty whoosh xapian-haystack > # pip3.5 install mailman mailman-hyperkitty For current Mailman the minimum Python version is 3.6 and all components are Python 3. Nothing is Python 2. You should be installing the latest versions of everything. Do not install Django explicitly with pip. It will be installed as a dependency of Postorius and Hyperkitty. If you are going to use Xapian, you don't need whoosh. Also, you need to install Xapian itself, either from source or with yum if there is a yum package. > And now Im in trouble, what would be the right steps from here. I'm not that competent to use pip, so im confused - should i install python36 packages and try to upgrade pip2 and pip3 mailman packages after that or should i install these pip2 packages with pip3 and remove from pip2 list? There are no pip2 packages. Everything is Python >=3.6. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

4 years, 1 month

[MM3-users] Re: [Django] ERROR (EXTERNAL IP): Internal Server Error: /hyperkitty/list/direttivo@catania.linux.it/top-threads

by Mark Sapiro

On 2/28/21 10:17 AM, blackout69 wrote: > Hello everybody,I received this error message and from that moment the > maiilan3 has stopped working.I restarted the following services mailman3 > and mailman3-web and everything started working fine.Can you tell me > what is the problem my django server has encountered? It looks to me like a race condition of some kind or possibly an incompletely deleted thread. Have you deleted any threads? I'm curious as to how you happen to be going to <https://lists.catania.linux.it/hyperkitty/list/direttivo@catania.linux.it/t…> as opposed to <https://lists.catania.linux.it/hyperkitty/list/direttivo@catania.linux.it/#…>. I know that ULS and similar ones are in HyperKitty's urls.py, but I don't see them exposed on any pages. Anyway, when you say "from that moment the maiilan3 has stopped working." (sic) exactly what stopped working. Did core stop delivering messages? Did Postorius stop working. If so, this all may be related to some underlying database issue or something like that that was fixed by the restarts. I don't see how this HyperKitty error could affect Postorius or Mailman core. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

4 years, 5 months

[MM3-users] Re: Strange messages after upgrading Mailman3

by Mark Sapiro

On 12/5/22 08:56, Allan Hansen wrote: > Thank you, Mark, as always. > > We upgraded to the latest versions: > > Mailman Core Version GNU Mailman 3.3.7 (Tom Sawyer) > Mailman Core API Version 3.1 > Mailman Core Python Version 3.9.2 (default, Feb 28, 2021, 17:03:44) [GCC 10.2.1 20210110] > > Also, I don't have any umbrella lists. I need more information about this. As far as I know, this can only happen with messages that are cross posted to multiple lists and held for moderator approval on more than one list. This should have been fixed in Mailman 3.3.6 by <https://gitlab.com/mailman/mailman/-/merge_requests/932>. This should never happen with a post that is delivered to members without being held. Also, I think when this happens to a delivered post, the post in the archive will also have the content lost message. I would like to get as much information as possible about the circumstances when this happens. In particular, was the post held for moderation, was it posted to multiple lists, for how many lists was the content lost. Also, it may help to see the full raw message with the content lost notice. (If there are privacy concerns you can send these to me off list) -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

2 years, 7 months

[MM3-users] Re: Replace "visit your dashboard" with URL link, like in Mailman 2.

by Mark Sapiro

On 5/8/23 09:52, Mark London via Mailman-users wrote: > Has anybody attempted to change the format of the emails that are sent > to the owner of a mailing list, to include a URL, instead of stating > "visit your dashboard"? The former is the behavior of Mailman 2. Many users, e.g. Jens Günther whose replied to this thread, have made custom templates to do this. > Is there any reason why this can't be done? If nobody has tried to do > so yet, I'll give it a try myself, as I'm familiar with django. Thanks. See https://gitlab.com/mailman/mailman/-/issues/901 and https://gitlab.com/mailman/mailman/-/issues/1055 Also see https://docs.mailman3.org/projects/mailman/en/latest/src/mailman/rest/docs/… for some documentation of templates, but note that you don't need to set URI's via REST, You can place list specific templates in Mailman's var/templates/lists/<list-id>/<lc>/ directory, domain specific templates in var/templates/domains/<domain-name>/<lc>/ and sitewide templates in var/templates/domain/site/<lc>/ or you can create list and/or domain specific templates via Postorius but see https://gitlab.com/mailman/postorius/-/issues/366 -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

2 years, 2 months

[MM3-users] Re: Fwd: Re: Removing a mail addresses and users

by Mark Sapiro

On 6/10/20 3:33 PM, Allan Hansen wrote: > > Some dedicated and skilled developers are currently investing heavily in a replacement for Postorius, but I really believe that any work is better spent transforming Mailman Core etc. into a true account-based system. Then, and only then, will it make sense to build new and better interfaces on top, as, without a healthy core, such interface will currently may have to inherit the confusion and disconnect that currently is so aggravating. I think you misunderstand what Mailman Core currently is. I think Mailman Core's concept of User is exactly what you are talking about when you ask for a "true account-based system". The problem is that Mailman Core was intentionally designed without any web UI and a RESTful API for separate web based UIs or whatever to use to communicate with it. Then the people who developed the web based management UI (Postorius) and archive UI (HyperKitty) chose to develop those within a Django framework and Django has its own concept of User separate from Mailman Core and that is where the disconnect occurs. It's not that Mailman Core lacks what you want. It's that Django doesn't use it. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

5 years, 1 month

[MM3-users] Re: Mailman3 don't work if aliases are enabled

by Mark Sapiro

On 10/9/20 3:37 PM, Massimo Zappalà wrote: > > If I try to uncomment the last two rows in main.cf > virtual_alias_maps = mysql:/etc/postfix/mysql_virtual_alias_emails.cf, mysql:/etc/postfix/mysql_virtual_alias_maps.cf > virtual_mailbox_maps = mysql:/etc/postfix/mysql_virtual_mailbox_maps.cf > The aliases and email messeges works fine but the mailman3's mailinglist don't work. > > How can I solve this? I'm not certain about this. The Postfix log entries for the failed delivery to Mailman may help, but I think you deed to configure an alias domain or domains for your list domain(s). The alias domain is not a real domain and is not in DNS. it is any unused domain name and will be known only to Postfix. I typically use x.example.com where example.com is the real list domain. You can set the alias domain for the real domain in the domains view in Postorius. Then you need to add `hash:/var/lib/mailman3/data/postfix_vmap` to your main.cf virtual_alias_maps setting and run Mailman's `mailman aliases` command to generate the proper /var/lib/mailman3/data/postfix_* files. See <https://docs.mailman3.org/projects/mailman/en/latest/src/mailman/docs/mta.h…>. You may also need to set relay_domains = hash:/var/lib/mailman3/data/postfix_domains -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

4 years, 9 months

[MM3-users] Re: Set domain inactive / archive import from other domain.

by Mark Sapiro

On 09/15/2018 07:15 AM, Johannes Winter wrote: > > we've migrated a list from another domain and would like to keep the > archives. Hyperkitty doesn't recognize the emails in the mbox archive > during import due to the other domain name. I don't understand the problem. I just ran a test where I imported a mbox containing 3 messages which were not even list posts and in which I had changed all the domain names to 'example.com'. I.e., in the mbox I imported, the headers containing domain names looked like From root(a)example.com Sun Aug 26 09:12:06 2018 Return-Path: <root(a)example.com> Delivered-To: root(a)example.com Received: by example.com (Postfix, from userid 0) From: Anacron <root(a)example.com> To: root(a)example.com Message-Id: <20180826161158.F154D34010C(a)example.com> X-msapiro-MailScanner-From: root(a)example.com and the 3 messages were all added to the archive. Why do you think the domain name is an issue? Perhaps the issue is there is already one or more messages in the archive and the messages you are importing are older. If you don't specify --since to hyperkitty_import it won't import messages older than the newest message in the existing archive. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

6 years, 10 months

[MM3-users] Re: non-member moderation: list default vs default processing

by Mark Sapiro

On 5/21/21 3:58 PM, dancab(a)caltech.edu wrote: > > Each time this problem has come up I've had to change the moderation setting from "List Default" to "Default Processing". > But I've also noticed that some non-members were imported and already have "Default Processing" configured. > > Can you help me understand what determines this setting in Mailman3? > > It seems this would be something addressed via the mailman import21 option I've been using to import the lists. It depends in part on the version of Mailman core as `mailman import21` has had some changes, but here's how it should work. Regexps in MM 2.1 *_these_nonmembers are imported into the corresponding MM 3 *_these_nonmembers. Addresses in MM 2.1 *_these_nonmembers are imported into MM 3 as non-members with their moderation_action set to `defer`, `hold`, 'reject` or `discard` according to which *_these_nonmembers they came from. `accept` maps to `defer` and the others to the same. Prior to Mailman 3.3.0, `accept` mapped to `accept` but `defer` is better. This can get very convoluted if the same address is both a member and in *_these_nonmembers, but in both MM 2.1 and MM 3, the member takes precedence. Is this not what you are seeing? -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

4 years, 2 months

[MM3-users] Re: How to store additional information about lists?

by Mark Sapiro

On 5/29/21 7:31 PM, Stephen J. Turnbull wrote: > > There won't be a membership adapter in Systers Mailman if there isn't > one in GNU Mailman, because they used the stock user database and > added fields to the User record to implement their membership > extensions. > > Membership adapter is a Mailman 2 feature that allowed use of an > external membership database (such as a corporate LDAP system). I > thought it was ported to Mailman 3 but maybe not. Actually, it's not quite that simple. Mailman 2.1 defines a MemberAdaptor class, the methods of which are used for everything to do with list members and memberships and are all documented in the base definition in Mailman/MemberAdaptor.py and all raise `NotImplementedError`. Also included in Mailman 2.1 is Mailman/OldStyleMemberships.py which subclasses MemberAdaptor and actually implements the methods, storing the data in the same config.pck pickle as the list settings. Various people have implemented various other MemberAdaptor classes for various purposes. https://bugs.launchpad.net/mailman/+bug/558106 contains several revisions of one using LDAP for the (read only) membership data. The Mailman 3 analog of this is the IMember interface defined in mailman/interfaces/member.py and implemented in mailman/model/member.py -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

4 years, 2 months

[MM3-users]Incorrect URLs

by Mark Sapiro

There are various URLs exposed in various emails from the list, many of which are wrong. The underlying issue is there are generic replacements such as $listinfo_uri and $user_options_uri in various templates such as the welcome.txt list welcome message and the footer-generic.txt message and digest footer. Now that the Mailman 3 core engine has been separated from the web UI, the core doesn't know what these replacements should be and there is not yet an interface to get them from Postorius or whatever the web UI is. In the mean time, I have made list specific versions of welcome.txt (done before announcing the list) and footer-generic.txt (just now) with the correct URLs hard coded. The purpose of this post is two-fold. To explain the issue, and to test the footer. There are probably more of these incorrect replacements and I will fix them as I notice them or they are pointed out to me. Note that I am aware that the https URLs in the List-Subscribe: and List-Unsubscribe: headers added to posts are wrong, but I haven't got a handle on this yet. The mailto: links in those headers are correct as are the Archived-At: and List-Archive: URLs as there is an interface to get those from the archiver. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

9 years, 3 months

[MM3-users] Re: What implications do GDPR have on Mailman, and how to requiring members to accept registration

by Mark Sapiro

On 5/17/18 3:35 AM, Henrik Rasmussen wrote: > > I am also concerned that new lists are created with a subscription policy "Confirm" as default, rather than at least "Moderate", which can cause people to be subscribed by others against there will. I'm confused. Confirm sends a confirmation request to the subscribed address which must be acted upon by the user before she is added to the list. Moderate just means approval by an owner/moderator which, it seems to me is more susceptible to a 3rd party subscribe succeeding - i.e., subscription requires membership in some organization. A subscribes B who is a member of the organization but doesn't want to be on the list. The moderator verifies that B is an organization member and approves the subscription. Granted, it may be possible for some brain-dead vacation responder to confirm a subscription request, but that's a separate issue. > Also, regardless of this setting, administrators can subscribe people without their knowledge. Granted that this can be abused, but not allowing admins to directly add members will probably create more problems than it solves. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

7 years, 2 months

[MM3-users] Re: mail stuck in "out" queue

by Mark Sapiro

On 6/13/20 6:30 PM, Kareem Zidane wrote: > Hi Mark, > > Sorry to bother you! Just wanted to loop back on the above to see if you might have any thoughts on that by chance? I'm sorry, I thought I had replied, but clearly I hadn't. So as I understand it, you have multiple lists in this Mailman 3 installation, and only one list has this issue, and mail to this list gets archived, but not delivered but is queued in var/queue/out and remains there. This is hard to understand. When Mailman's out runner processes its queue, it processes messages in chronological order without regard to which list they are from, so it must be at least attempting to deliver these messages, and somehow failing and requeueing them. I do see this Jun 09 17:36:31 2020 (82) Cannot connect to SMTP server postfix on port 25 in the log excerpt you posted. What's in Mailman's smtp.log and the postfix mail.log that might be related to this? It's hard to understand how an MTA connect issue could affect only one list. If it is accepting the connict and then dropping the connection for too many invalid RCPT addresses, that's a possibility. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

5 years, 1 month

[MM3-users] Re: Mailserver Log entries following through mailman?

by Mark Sapiro

On 11/10/23 12:39, Andreas Barth wrote: > > is there some way to find an ID to follow an email through mailman? In > the exim logfile there is only something like C="250 Ok" (whereas in > mails sent to another exim there is something like C="250 OK id=…"). > (What I sometimes need to do is to follow an mail whether it is sent > through mailman and to where.) Every message should have a unique Message-ID: header. According to my reading of https://www.exim.org/exim-html-current/doc/html/spec_html/ch-log_files.html Exim should create a => log entry with an id= field containing the Message-ID when the message is received for delivery to Mailman and another similar entry when the message is received from Mailman for delivery to one or more list members. Mailman itself logs ACCEPT:, HOLD:, DISCARD:, REJECT: and OWNER: entries in mailman.log for those events (OWNER: indicates a message to list owners). All these entries contain the Message-ID. Mailman's smtp.log contains 'post to' and 'smtp to' entries with Message-IDs for all outgoing messages. > Or should I raise an issue as whishlist request? What exactly would you wish for that Mailman can do. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

1 year, 8 months

[MM3-users] Re: Mailman3 on AlmaLinux 9 - Python3.9 errors

by Mark Sapiro

On 7/1/24 12:34, willie.castillo(a)caemilusa.com wrote: > Hi, I'm trying to configure mailman3 on my AlmaLinux 9 server and when checking the mailman3 status, I get the following: ... > File "<string>", line 8, in run_migrations > File "/usr/lib/python3.9/site-packages/alembic/runtime/environment.py", line 851, in run_migrations > self.get_context().run_migrations(**kw) > File "/usr/lib/python3.9/site-packages/alembic/runtime/migration.py", line 608, in run_migrations > for step in self._migrations_fn(heads, self): > File "/usr/lib/python3.9/site-packages/alembic/command.py", line 309, in upgrade > return script._upgrade_revs(revision, rev) > File "/usr/lib/python3.9/site-packages/alembic/script/base.py", line 435, in _upgrade_revs > return [ > File "/usr/lib64/python3.9/contextlib.py", line 137, in __exit__ > self.gen.throw(typ, value, traceback) > File "/usr/lib/python3.9/site-packages/alembic/script/base.py", line 259, in _catch_revision_errors > raise util.CommandError(resolution) from re > alembic.util.exc.CommandError: Can't locate revision identified by '2156fc3f6f7d' > > > Any ideas why? It looks like your mailman/database/alembic/versions/ directory is missing 2156fc3f6f7d_add_dmarc_addresses.py. What is your mailman version? This migration was added in Mailman core 3.3.9. Possibly your Mailman version is < 3.3.9 but the alembic_version table in your mailman database points to an alembic version not in your mailman core version. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

1 year, 1 month

[MM3-users] Re: 501 Message has defects

by Mark Sapiro

On 10/25/24 08:06, jgn1980--- via Mailman-users wrote: > Hi > > I'm using GNU Mailman 3.3.8 (Tom Sawyer) on debian 12, and I have this issue on a scheduled e-mail, every day: > > 2024-10-25T04:00:04.874081+01:00 XXXXXXX postfix/lmtp[29258]: CEB32100F44: to=<XXX(a)XXX.XX>, relay=localhost[127.0.0.1]:8024, delay=0.02, delays=0.01/0/0.01/0.01, dsn=5.0.0, status=bounced (host localhost[127.0.0.1] said: 501 Message has defects (in reply to end of DATA command)) Mailman's LMTP runner is parsing the incoming message and it has defects. To see what the defects are, put the raw message in a file and do the following in `mailman shell` ``` $ mailman shell Welcome to the GNU Mailman shell Use commit() to commit changes. Use abort() to discard changes since the last commit. Exit with ctrl+D does an implicit commit() but exit() does not. >>> import email >>> from mailman.email.message import Message >>> msg = email.message_from_binary_file(open('test.eml', 'rb'), Message) >>> msg.defects ``` This should display the defects. Once you know what the defects are, you may be able to adjust the process that produces the scheduled email to create a message without defects. Also see https://gitlab.com/mailman/mailman/-/issues/351 which we should fix to report and/or log the defects. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

9 months, 1 week

[MM3-users] Re: mail stuck in "out" queue

by Mark Sapiro

On 12/17/24 13:38, pdw(a)mythic-beasts.com wrote: > > I've now found the SHUNT entries and the exceptions seem related to restarting mailman (see below) so that much makes sense. > > unshunting does indeed restore them, but unfortunately they're now back in the looping state. This behavior always seems related to decorate being unable to get templates for for the message header/footer. Does this list define templates for list.member.regular.header or list.member.regular.footer in Postorius? To rule out a template retrieval issue as the cause, you could try this patch to skip decorating the message. ``` --- a/src/mailman/handlers/decorate.py +++ b/src/mailman/handlers/decorate.py @@ -40,6 +40,7 @@ alog = logging.getLogger('mailman.archiver') def process(mlist, msg, msgdata): """Decorate the message with headers and footers.""" + return # Digests and Mailman-craft messages should not get additional headers. if msgdata.get('isdigest') or msgdata.get('nodecorate'): return ``` If you apply that patch and the messages get delivered, albeit without the footer, the issue is definitely in retrieval of templates. You say you have removed the var/templates/list/LIST_ID/* templates, but is there still a var/templates/list/ and a var/templates/list/LIST_ID/ directory and are they searchable by the Mailman user? -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

7 months, 2 weeks

[MM3-users] Re: [Django] ERROR (EXTERNAL IP): Service Unavailable

by Mark Sapiro

On 2/22/21 5:47 PM, dancab(a)caltech.edu wrote: > I've imported a list from Mailman2 and had the owner send a message for testing purposes. > > His message are moderated per the list configuration. The message is delivered to Mailman3 and shows that it's being held. > > However if we attempt to view the message in Postorius web UI then an error is thrown. Can you see the message in the held messages view and only get the error when you click the message, or can you not see the held messages view at all? Is it only this list that has an issue? Can you visit other pages for this list? > In the web interface the error is simply "Mailman REST API not available. Please start Mailman core." > > But Mailman Core is running and we can do other functions just fine. Our superuser account receives a much more detailed message with the subject below. > > [Django] ERROR (EXTERNAL IP): Service Unavailable: /postorius/lists/cpa-announcement.caltech.edu/held_messages > > Does "Service Unavailable" mean I'm missing something in the configuration? Possibly, but I don't think so. What is the Traceback from the message Django sends to the server admin? -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

4 years, 5 months

[MM3-users] Re: mailman 2.1 or 3.0 (or another MLM)

by Mark Sapiro

On 1/26/22 08:49, nate wrote: > (In case the list admin/moderator sees this message I sent this email > again from the wrong FROM: address a few minutes ago, re-sending with > the right one now, please discard that bad message(have not > gotten confirmation it was held for moderation) Posts to this list from non-members are discarded without notice. > -- side note with > MM2 it gave the user the option to discard their own message if it > was held for moderation if I recall right. My very first post to > this list was held for moderation (sent from correct FROM: address), > not sure why it was held(it got through eventually), was assuming > that perhaps I triggered a bug in that the fact I am using the > text mailman in my email address maybe tripped something I don't know, > but I was seemingly unable to view the status/discard that message.) Your initial post was held because all new members are moderated by default. This is an anti-spam measure. If you post on topic, we clear your moderation. And yes, Mailman 3 doesn't (yet?) have a way for ordinary users to delete their own held posts. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

3 years, 6 months

[MM3-users] Re: MASS mailing list import - 900 ezmlm lists to Mailman

by Mark Sapiro

On 4/18/23 17:41, grant(a)horizonsunlimited.com wrote: > Hi, > I hope someone can help. > I have 900-ish, mostly small lists, basically all locations round the globe where people have volunteered to help particular people passing though. Those lists have 2 to 200 people. There are about 50 lists of up to 1000 people each we use for announcements. So two different setups, and a whole lot of manual work. > I am moving from an old, now unreliable, host to a new one. Hopefully reliable, certainly cheaper, and better for privacy GDPR etc regs. > Mailman is already installed, appears to be version 82?? on cPanel 108.0.15. Ezmlm, no surprise, is not. Installing and configuring it is not something trivial, also have to install gmail etc. Yuck. First of all, cPanel Mailman is most likely 2.1.38. This list is for Mailman 3. The appropriate list for Mailman 2.1 is mailman-users(a)python.org <https://mail.python.org/mailman3/lists/mailman-users.python.org/>. This search <https://mail.python.org/archives/search?mlist=mailman-users%40python.org&q=…> finds 169 posts mentioning ezmlm. Some of those might be relevant to importing lists from ezmlm. Also see <https://wiki.list.org/DOC/Mailman%20and%20CPanel> and be aware that Mailman 2.1 is end of life. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

2 years, 3 months

[MM3-users] Re: Mailing list welcome message won't update

by Mark Sapiro

On 1/16/25 13:43, via Mailman-users wrote: > HI Mark, > > I simply tested by unsubsribing to the mailing list and then re-subscribing to receive the welcome message. > > How do you test using the mailman command with user list? sudo -u list mailman addmembers -h or in your case, you may not have that subcommand in which case see sudo -u list mailman members -h However, that is all irrelevant. I only thought you might be testing by using `mailman` commands to add a member and if so and if the command is not run as user `list`, it might not have access to read the template. If you are (un)subscribing via Postorius or via email these are handled by Mailman runner processes which should all be running as user `list` and should have access to the template. So, as I said, I don't know what the issue is unless it is something with your Debian package which is ancient. Even Debian buster has mailman core 3.2.1. Does your empty list:member:generic:footer.txt work to not add any footer to list messages? If so, I can't think of any reason why your list:user:notice:welcome.txt wouldn't be used. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

6 months, 2 weeks

[MM3-users] Re: The MAILMAN_ARCHIVER_KEY was not sent as the Authorization HTTP header.

by Mark Sapiro

On 10/22/21 12:53 PM, Marco Marinello wrote: > Dear Mark, > > had > # pip freeze > ... > HyperKitty==1.3.5 > ... > mailman-hyperkitty==1.2.0 > ... > > and got the described behaviour. > Downgraded to > > HyperKitty @ https://gitlab.com/mailman/hyperkitty/-/archive/1.3.4/hyperkitty-1.3.4.zip > mailman-hyperkitty @ https://gitlab.com/mailman/mailman-hyperkitty/-/archive/v1.1.0/mailman-hype… > > and now archiving works. Actually, with HyperKitty==1.3.5 and mailman-hyperkitty <1.2.0, the message is not exactly as in this message's Subject:. It is 'The MAILMAN_ARCHIVER_KEY is now required to be sent over the Authorization HTTP header, you need to upgrade the mailman-hyperkitty package to 1.2.0 or newer. Perhaps you paraphrased the message. If not, that exact message was not in any version of HyperKitty leading up to 1.3.5. And with mailman-hyperkitty==1.2.0, it should work. I see what pip freeze reported, but something was wrong somewhere because that combination should work and the 'The MAILMAN_ARCHIVER_KEY ...' message would only be logged with a version of mailman-hyperkitty <1.2.0. Perhaps you did not realize you had to restart Mailman core and skipped that step. The mailman-hyperkitty module is imported by Mailman core, thus core needs to be restarted if it is changed. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

3 years, 9 months

[MM3-users] Re: Threads by month dropdown

by Mark Sapiro

On 6/5/25 10:22, Sam Darwin via Mailman-users wrote: > > Deleting all archives and importing the older messages first... that worked correctly. It is a large destructive actions though, should that count as the official solution? No, that shouldn't be the "official" solution. It may be too late now, but did you restart mailman-web? See the thread (started by you) at https://lists.mailman3.org/archives/list/mailman-users@mailman3.org/thread/… > Detour: > > While debugging it seemed to be an interesting question whether the mailman3-web periodic cron tasks had been running, and if those tasks affected the hyperkitty import. Referring to "Cron Jobs for Mailman Web" https://docs.mailman3.org/en/latest/install/virtualenv.html#cron-jobs-for-m… Those instructions are not dumping stdout/stderr to a log file, nor rotating the log files. What is your opinion if I sent a pull request that adds "Method 2 - Advanced", and include logging for the crons. Or maybe I am missing something. Is the output of the cron tasks getting logged in your environment? Any output to stdout or stderr from any of the crons will be emailed to the owner of the crontab or the address in a MAILTO= in the crontab. You should ensure that mail to `mailman` is deliverable -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

1 month, 3 weeks

[MM3-users] Re: Starting new thread in Postorius -> disappears

by Mark Sapiro

On 11/21/2017 01:53 AM, Thor Atle Rustad wrote: > I have had this working before, so I guess it's a simple setup error. > > When I send a message from the Postorius interface, it is neither sent out > or stored. I briefly see this message: "Your reply has been sent and is > being processed.", and then nothing happens. I assume you mean HyperKitty. Postorius has no facility for posting messages to a list. > I have installed maxking's docker-mailman and copied my old setup with > basic stuff like passwords and such. The only new thing is that I'm now > running in userspace by putting in /etc/subuid and subguid: When you post from HyperKitty, HyperKitty creates a django.core.mail.EmailMessage() instance and calls it's send() method which causes Django to send the message to the list address. It then has to be received and processed by Mailman the same as any email post before it will appear in the archive. Are you running with DEBUG = True. If so, Hyperkitty won't send the mail. If DEBUG = False and normal posting is working, perhaps there is some Postfix issue in delivery of the message from Django to Mailman. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

7 years, 8 months

[MM3-users] Re: Mailman runners missing - nothing in logs

by Mark Sapiro

On 1/18/20 6:13 AM, Torge Riedel wrote: > > So now back to good. I checked all the logs prior to restarting the > service in directory /opt/mailman/var/log, but I do not see any errors. > And the master was still running, I would expect that the master will > start / restart a runner in case it died? What's in /opt/mailman/var/log/mailman.log? There should be entries like Jan 18 13:51:11 2020 (11700) xxxx runner started. for each runner every time it starts and like Jan 18 13:47:31 2020 (10096) xxxx runner caught SIGTERM. Stopping. Jan 18 13:47:31 2020 (10096) xxxx runner exiting. for each runner every time it stops, but with perhaps a reason other than `caught SIGTERM` That should at least tell you when they stopped. Also, if they stopped for some reason. The master will only restart runners that exit if they exit because of SIGUSR1 or some internal error and even then, only the configured max_restarts number of times. Do you have a logrotate script for the logs in /opt/mailman/var/log, and if so, does it have a postrotate script that signals Mailman to reopen logs other than by a `mailman reopen` command, perhaps by signaling the master with other than SIGHUP? -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

5 years, 6 months

[MM3-users] Re: Bounce counter visibility for List Admins

by Mark Sapiro

On 5/19/21 4:10 AM, smorytko(a)gmail.com wrote: > As you may know, we use MM3 but the Affinity/Empathy flavor and we only have List Administrator using the Postorious/Affinity. Unlike many here, we have no access to any server-based files. > > Can someone tell me if Postorious gives list administrators visibility into the bounce counters? User's bounce scores are not available in REST so neither Affinity nor Postorius has access to them. > While List Admins get to set various bounce counters we have no visibility of those counters until, all of a sudden, an entire source of subscribers hits the bounce threshold and gets bounced off the list. It would be really nice to be able to see the bounce counters and possibly head off trouble before users/subscriber mail hosts. It can serve as an effective early warning system that something is wrong. > > Assuming Postorious cannot, is there another way List Admins can inspect bounce counters? Email command? In the next release of Mailman core (3.3.5), lists will have a `bounce_notify_owner_on_bounce_increment` setting which if True, will send to the list admins a notice including the triggering DSN when a user's bounce score is incremented. This setting will also be exposed in Postorius 1.3.5. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

4 years, 2 months

[MM3-users] Re: [Mailman-Developers] Mailman 3.1 beta coming soon

by Mark Sapiro

On 12/02/2016 08:07 AM, Barry Warsaw wrote: > Hello Mailman 3 users and developers! > > It's been a long slog but I think we are finally ready to begin the release > process for Mailman 3.1. I am pretty happy about the state of the Core, and > Florian is going to do some more integration testing with HyperKitty and > Postorius. We still have a few issues and a merge proposal or two to address, > but I am planning to spin a beta of Core this weekend to facilitate testing. Barry, I understand you need to feel comfortable about everything going into the core and this takes time and energy to review all the MRs, but I think it is pretty important to get the DMARC mitigations into the core. If there's anything I can do to help with this, let me know. I will rebase my branch again and possibly make one minor tweak, but I think it should be OK. I can also work on Postorius. Also in this vein I have a question. Should the DMARC settings be added to the Alter Messages tab in the List Settings view or would it be better to have a separate DMARC Mitigations tab. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

8 years, 7 months

[MM3-users] Re: hyperkitty_import Unicode Errors

by Mark Sapiro

On 5/6/23 11:27, Dave Hall via Mailman-users wrote: > Hello. > > When running hyperkitty_import to import Mailman 2.1 archives, I frequently > get errors like > > django.db.utils.OperationalError: (1366, "Incorrect string value: > '\\xE1\\x90\\xA7\\x0A\\x0AO...' for column > `mailman3web`.`hyperkitty_email`.`content` at row 1") > > Is there a known cause for this? Is it fixed in some release more recent > than 0+20180916? I'm not sure what's going on here. My first thought is the database is MySQL or MariaDB and this is a 4-byte UTF-8 encoding and the database column charset definition is utf8 and not utf8mb4, but '\\xE1\\x90\\xA7\\x0A\\x0A' is the 3-byte UTF-8 encoding for CANADIAN SYLLABICS FINAL MIDDLE DOT ('\\xE1\\x90\\xA7') followed by two newlines. so I'm not sure what the issue is. When you encounter this error, there should also be a message like "Message <Message-ID> failed to import, skipping" or perhaps "Failed adding message <Message-ID>: <error message>". Using that Message-ID value, you should be able to find the original message in the input mbox. I would like to see that mbox message. You can remove any personal info. I only really need the message body and any Content-*: headers. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

2 years, 2 months

[MM3-users] Re: MM3 Postfix FROM header issue

by Mark Sapiro

On 12/17/23 1:20 PM, eboltz(a)lhtservices.com wrote: > Okay, created a new Mail Host with only 'list.example.com' and no alias domain. I then created a new list and tried subscribe/unsubscribe but got the following error message in mail.log file > > Dec 17 21:14:26 mailman3 postfix/smtpd[6649]: 057E744552: client=localhost[127.0.0.1] > Dec 17 21:14:26 mailman3 postfix/cleanup[6653]: 057E744552: message-id=<170284766462.1144.7361224077829393518(a)ec2-54-172-107-21.compute-1.amazonaws.com> > Dec 17 21:14:26 mailman3 postfix/qmgr[6644]: 057E744552: from=<list_test-bounces(a)lists.example.com>, size=908, nrcpt=1 (queue active) > Dec 17 21:14:26 mailman3 postfix/smtpd[6649]: disconnect from localhost[127.0.0.1] ehlo=1 mail=1 rcpt=1 data=1 quit=1 commands=5 > Dec 17 21:14:27 mailman3 postfix/smtp[6654]: 057E744552: to=<eboltz(a)lhtservices.com>, relay=mail.smtp2go.com[45.79.114.202]:587, delay=1, delays=0.01/0.01/0.52/0.46, dsn=5.0.0, status=bounced (host mail.smtp2go.com[45.79.114.202] said: 550-From header sender domain not verified (lists.example.com) 550-On your Sending > Verified Senders page 550 verify the sender domain or email to be allowed to send. (in reply to end of DATA command)) Your relay host doesn't like lists.example.com as a sender. You need to fix this on the relay host mail.smtp2go.com. It seems to say that you need to go to the Sending > Verified Senders page on the relay host and add lists.example.com. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

1 year, 7 months

[MM3-users] Re: Changes in Django admin interface are not syncing to postorius gui /database

by Mark Sapiro

On 11/11/24 04:17, Ulf R. wrote: > Now remains one question: We have a (internal) Active Directory (AD) Authentication for Mailman3. > So everyone who wants to work as an owner nor moderator must be an AD-User. > In hyperkitty I am able to assign these list rights directly to AD-Loginnames. > With the logic above I have to assign these rights in postorius too by assining the mail address of the users AD-Account? (I see no AD-Data here). You are conflating Postorius, which is simply a web management UI for Mailman core, with Mailman core itself. I suppose it's not really important, but when you think of access rights in Postorius, these are actually in mailman core. > Or would it works to assing the mail address of the users AD-Account first an then the allocation will automaticly synced to hyperkitty - maybe using the mail address as key? > What would be the best practice for owner/moderator user management? Just ensure that list owners and moderators are using the email address of their AD-Account. This will propagate to HyperKitty. You shouldn't need to manipulate HyperKitty access rights directly at all. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

8 months, 2 weeks

[MM3-users] Re: Getting an error for every post made to any list

by Mark Sapiro

On 4/22/25 1:35 PM, sun zi via Mailman-users wrote: > As mentioned in the subject, I am getting the following error after every post. Any assistance would be much appreciated: The error is from an attempt to archive a message. Since it fails, the problem message is queued in Mailman's var/archives/hyperkitty/spool/ directory and is retried with each post. The message is in a Python pickle, *.pck, file in that directory. Yoy can examine the message by giving it to `mailman qfile`. From the tracebacks you posted, it appears that the issue is that the sender's address is longer than 255 bytes. Move the .pck file out of the var/archives/hyperkitty/spool/ directory and that will stop the recurring error. If the message is spam you can ignore it. If you want to archive it you need to edit the message's sender, but you can't edit the *.pck directly. You can use `mailman shell` and do something like ``` import pickle with open('path/to/the/pickle', 'rb') as fp: msg = pickle.load(fp) # code here to edit the sender with open('path/to/the/pickle', 'wb') as fp: msg = pickle.dump(msg, fp) ``` -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

3 months, 1 week

[MM3-users] Re: very few moderation capabilities for moderators

by Mark Sapiro

On 9/18/21 1:41 AM, Guillermo Hernandez (Oldno7) via Mailman-users wrote: > > I've been trying to "translate" the python script that I made for > mailman2 to new mailman3, but obviously, I'm missing something. It seems > that mailman cannot locate the "module" needed. > > The script is named "moderar.py": > > *-*-*-*-*-*-*-*-*-* (begin of /usr/local/mailman3/moderar.py) > > /usr/local/mailman3 # cat moderar.py > #! /usr/local/bin/python3 > # > # script via withlist para moderar a pelo en mailman3 > # > > import os > import sys > > sys.path.insert(0, '/usr/local/mailman3') If /usr/local/mailman3 is not already in the path, this is too late. The rest of your script is not relevant. Here's what you need: ---------------------------------------------------- from mailman.interfaces.action import Action def moderar(mlist, member): memb = mlist.members.get_member(member) if not memb: print('{} is not a list member.'.format(member)) return memb.moderation_action = Action.hold print('{} moderation_action set to hold.'.format(member)) ---------------------------------------------------- That should be put in a moderar.py file somewhere in Mailman's path. The same directory that contains the `mailman` command is for sure good. Possibly that's /usr/local/mailman3/bin/. Then mailman withlist -r moderar -l list_to_use(a)mydomain.com address(a)moderate.com should work. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

3 years, 10 months

[MM3-users] Re: Internal Server Error: Reattach this thread

by Mark Sapiro

On 10/18/22 21:38, hasimoto(a)designet.co.jp wrote: > Hi, > > In the archive, clicking "Reconnect this thread" causes an internal server error. > I received the following email, but I don't know how to solve it, please help me. How big is the archive in which you are doing this? Are you running the periodic django-admin jobs via cron? See https://gitlab.com/mailman/hyperkitty/-/blob/master/example_project/crontab I ask because I've now seen this issue in a test installation, but only with an archive containing only one message. The particular steps I followed somewhat by accident were: Start with an archive containing only two messages in separate threads. Open one thread and `Delete this thread` leaving only one thread. Open that thread and click `Reattach this thread` - this produces the error. Restore the deleted message/thread with `hyperkitty_import` Clicking `Reattach this thread` in either thread produces the error. After running the hourly and daily django-admin jobs, clicking `Reattach this thread` in either thread works as expected. That scenario with deleting and restoring a thread was somewhat accidental, but did produce the error. in other list archives on the same server, even one containing only one message, clicking `Reattach this thread` in either thread works as expected. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

2 years, 9 months

[MM3-users] Re: certain msgs not archived

by Mark Sapiro

On 2/17/22 10:49, Sandor Dibuz wrote: > > Great! So valuable info. Thank you, Mark. "earlier that the oldest > message" you sure mean earlier then the oldest of already archived. Or > since= any date of the mbox messages to be imported. Will check the > hyperkitty_import for since= option format. That was a typo. I meant "earlier than". I tend to use `since=1970-1-1`. It doesn't matter how early as long as it's earlier than the oldest message to be imported. > Yet another question. Is it supposed to stop minutely/hourly/etc. > mailman3-web cron jobs during import to avoid any collisions? Or even > mailman3-web completely but probably it needs to be running. I only ask > because my last import attempt on a test maillist failed with > > [ERROR/MainProcess] Failed indexing 1 - 2 (retry 5/5): (1712, 'Index > hyperkitty_attachment is corrupted') (pid 5841): (1712, 'Index > hyperkitty_attachment is corrupted') > > Optimizing cured the hyperkitty_attachment InnoDB index corruption > problem but scared the sh.. out of me for a moment... I don't know what may have caused the index corruption. I don't bother to stop any crons or services while hyperkitty_import is running and I've never had an issue. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

3 years, 5 months

[MM3-users] Re: Mailman3 on virtual domains

by Mark Sapiro

On 10/7/23 09:25, Lars Bjørndal wrote: > [Mark] > > [...] > >> See https://docs.mailman3.org/projects/mailman/en/latest/src/mailman/docs/mta.h… > > Ok, so I have to create an alias_domain in Mailman. Is it possible to do that from the web interace? It has been in postorius since 1.2.0. > I find an Edit button for each domain, and then I may change name and description. But I can't find an option to add an alias_domain. As I'm blind, and using a text only browser, some information may have been missing. Or can I do it from the command line? If your Postorius is >=1.2.0, there should be 3 fields on the Edit domain page - Description, Alias Domain and Web Host. The whole page looks something like: Edit domain example.com Description Alias Domain Normally empty. Used only for unusual Postfix configs. Web Host The domain from which you want the web UI to be served from. This can be same or different from the Mail Host. You can edit the list of available web hosts here. You can also set an alias_domain in `mailman shell`. E.g., ``` >>> domain = getUtility(IDomainManager).get('example.com') >>> domain.alias_domain = 'x.example.com' >>> commit() ``` -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

1 year, 9 months

[MM3-users] Re: ARC signing outgoing mails

by Mark Sapiro

On 11/1/19 4:04 PM, Wolfgang Bock via Mailman-users wrote: > > Hello, > > in > https://mailman.readthedocs.io/en/latest/src/mailman/handlers/docs/arc_sign… > I found this hint > > "It is highly recommended that Mailman maintainers configure ARC siging > of their outgoing email. ..." > > So I added in my mailman.cfg: > " > [ARC] > privkey: /path/to/private.key > " > > > But I don't understand the meaning of > > > "... and located at: > " > #{config.ARC.selector}._domainkeys.#{config.ARC.domain} –ex.– > test._domainkeys.example.com > " This doesn't seem correct. The DNS TXT record should be added to #{config.ARC.selector}._domainkey.#{config.ARC.domain} (note _domainkey, not _domainkeys) I.e., this is a normal DNS TXT record containing the public key for DKIM signing. > Where can I find this string or where do I have to place it. I am using > bind9 as dns-handler. Read the entire ARC section beginning at <https://gitlab.com/mailman/mailman/blob/master/src/mailman/config/schema.cf…>. You have to configure more than just privkey. Once you have configured enabled: yes privkey: /some/location/yourprivatekey.pem selector: test (or whatever you want) domain: your_domain.com The substitutions for #{config.ARC.selector} and #{config.ARC.domain} should be more clear. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

5 years, 9 months

[MM3-users] Re: Must every Moderator/Admin be member of a list to be able to use Postorius?

by Mark Sapiro

On 1/31/23 10:25, mm3-omf(a)sh.werbittewas.de wrote: > we're using debian/devuan with distribution-mailman3 and tried to add > admin/mod for a list, but if logging in as this user, there's no item to > manage the list at all. That's strange. If the logged in user is a moderator she should see the Subscription requests and Held messages tabs and with Postorius >= 1.3.9b1 also the Members tab, and if she is an owner she should see the full admin UI without needing to be a member. Are you sure the logged in user is the same user that's an owner/moderator? Log in as a superuser and find the user in the User's view fron the top nav bar in Postorius, click Manage to see all that user's roles and web account. > do we have to add this user as member of this list? > (that wouldn't be nice, because the user doesn't want to get normal > Mails from List, but only admin-Mails) It shouldn't be necessary, but a delivery can always be disabled for a member. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

2 years, 6 months

[MM3-users] Re: Sending a message to a subscription applicant when a subscription request is made

by Mark Sapiro

On 3/28/23 09:19, Allan Hansen wrote: > Mark, > > I tried doing this. > I unsubscribed myself from a list, set the policy as you suggest and re-applied. > I got no confirmation messages but as moderator I got the moderate message. How did you reapply? with the subscribe policy confirm then moderate a new subscriber should get the list:user:action:subscribe message and have to confirm before the request waits moderation? However, see below. > I then tried a bogus email and saw that a message was sent to that and no moderator message to me. > Would the server skip the verification message because I'm a moderator/admin? It's not because you are an admin. It's because you were logged in when you subscribed from the list's info page. You will need to confirm if you subscribe anonymously without logging in or subscribe via email, but if you are logged in to Postorius when you request subscription, that is sufficient confirmation that the request really comes from you, so you don't get an email confirmation. So, depending on your circumstances, this may or may not work for you. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

2 years, 4 months

[MM3-users] Re: "Cannot connect to SMTP server localhost on port 25"

by Mark Sapiro

On 9/22/19 2:11 AM, Manos Pitsidianakis wrote: > > mailman3 tries to get the template I've set via the mailman suite > interface but it assumes it is hosted locally: > 'http://localhost:8000/mailman3/api/templates/list/meli-announce.meli.delive…' > In my setup, they are separate machines. I can't find any way to change > this in the configuration. Also, can't the same template be loaded from > the master's localhost:8001 api? See <https://mailman.readthedocs.io/en/latest/build/lib/mailman/rest/docs/templa…>. There are other ways to set the URI for a template which will work in your case. When you set a template via Postorius, it sets the URI for that template to point to itself to return what it set. In your case, this URI points to localhost which is Postorius' localhost, but it is set in core and core's localhost is a different machine so it doesn't work. You could file an issue at <https://gitlab.com/mailman/postorius/issues> about this. > Haven't fiddled with python for some time, why is this exception catched > by "except socket.error"? Is it a subclass of > requests.exceptions.ConnectionError? The other way around. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

5 years, 10 months

[MM3-users] Re: Setting for archiving multi email domains?

by Mark Sapiro

On 3/9/21 4:05 PM, Duane Raymond wrote: >> As I said there, Mailman is going through the roster of regular >> (non-digest) list members to find those with delivery enabled and it >> finds a member that has no associated address. Can you view the members >> of this list? > > Yes - the list has three members (test list): me and two others. For my subscription I have gone through the membership options and explicitly set them. For the others their membership options are all unselected and thus I assume defaulting to a fallback value. (not sure if that matters) What matters is at least one of those members doesn't have an associated address record. If you have access to `mailman shell` do the following: mailman shell -l LIST_ID Welcome to the GNU Mailman shell Use commit() to commit changes. Use abort() to discard changes since the last commit. Exit with ctrl+D does an implicit commit() but exit() does not. The variable 'm' is the LIST_ID mailing list >>> for mbr in m.members.members: ... if mbr.address is None: ... print('user: {}, user.addresses: {}'.format( ''' mbr.user, mbr.user.addresses)) ... And see what that gives you. (At the last ... prompt just type `enter`) -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

4 years, 4 months

[MM3-users] Re: No digest.mmdf

by Mark Sapiro

On 3/28/21 8:18 AM, Christine wrote: > I have mailman3 running now, more or less. Mailman did not send messages > to my list, giving a permiission error on the digest.mmdf file, which > was not there. I created the file and made list.list the owner, now it > sends messages. > I don't know what should have created the file or what set the > permissions to the directory, I just thought I'd let you know. Mailman core creates and writes the lists/LIST-ID/digest.mmdf file to accumulate messages for the digest. When a digest is sent the file is removed only to be recreated on the next post, and you will probably run into this issue again. The issue is the user that Mailman core is currently running as does not have permission to create that file. This usually results from having run core as root at some point in the past causing some files and directories to have been created by root and not writable by the mailman user now running core. You need to ensure that Mailman processes are always run as the Mailman user, often `mailman` but in the case of Debian/Ubuntu packages, `list`. You probably need to ``` sudo chown xxx:xxx path_to_mailman_var/ ``` where xxx is the Mailman user. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

4 years, 4 months

[MM3-users] Re: "autorespond to list owner" setting when importing from mailman2 to mailman3

by Mark Sapiro

On 2/25/22 08:32, bob B via Mailman-users wrote: > In mailman2 a list was moderated, the admin would also get an email to approve the message, our lists mainly have admins and no moderators. The admins did the moderation. > > In mailman3 it looks this is now a setting "autorespond to list owner" was this a setting in mailman 2? Yes, in MM 2.1 this was in the Auto-responder section of the admin UI, but this isn't what you wan. > Anyway, in importing email list from mailman 2 to 3, it appears that this setting is set to "No automatic response", is that right? Probably. It is =correct if that was the MM 2.1 setting. > So does this means if I import all my lists from mailman 2, then the admins will not get moderation notifications until I make that setting change on each list? No. The setting you want was Mailman 2.1's General Options -> admin_immed_notify which should be correctly imported to what Postorius shows as Settings -> Automatic Responses -> Admin immed notify and whether that is Yes or No, they should get periodic notices as long as cron is running `mailman notify`. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

3 years, 5 months

[MM3-users]Re: Sass and where to find core rest logs

by Mark Sapiro

On 05/04/2016 09:45 AM, Andrew Hodgson wrote: > Anyway, I have compiled Mailman3, and the web interface is coming up. Bringing up the list of lists initially gave me an exception, and I saw a 5xx error in the mailman.log file in /opt/mailman/var/logs. I read up on this and as a result downgraded Falcon<1.0 in the Python 3.4 venv. > > I am still unable to browse the list of lists with a 500 error, only this time I am not getting an entry in the mailman.log indicating there is an issue, and I am wondering if there are more detailed logs that can be switched on? There are the web server logs and there is the mailman-web configured in LOGGING:handlers:file in (using bundler) mailman-bundler/mailman-web/production.py (or testing.py). The production default is /var/log/mailman-web/mailman-web.log; the testing default is /opt/mailman/var/mailman-web/hyperkitty.log. > The other question I have is on Sassc. I believe the default Mailman web interface has now switched to the Ruby implementation of Sass, but is this correct? The documentation on the Mailman-Bundler site seems to suggest that Sassc is still the default. I think the bundler doc is out of date on this. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

9 years, 2 months

[MM3-users] Re: How to promote non-members to members?

by Mark Sapiro

On 3/17/22 12:25, Quanah Gibson-Mount wrote: > > I've done nothing in regard to this user. Neither command returns > anything. Well, something has deleted the user, but ... >> If the user is gone, you should be able to replace >> ``` >> ... sub_mgr.confirm(token) >> ``` >> in the above with >> ``` >> ... wfm = getUtility(IWorkflowStateManager) >> ... pendings.confirm(token, expunge=True) >> ... wfm.discard(token) >> ``` >> to clear the pending subscription and workflow and then be able to mass >> subscribe the user. > > I did this, and can confirm that they no longer show up in the list of > pending subscriptions via the mailman shell interface for that list. > > However, if I go to mass subscribe them via the interface, I get: > > HTTP Error 409: b'Subscription request already pending' I don't understand. If there is no entry for that address when you do ``` >>> for token, data in pendings.find(pend_type='subscription'): ... if data['token_owner'] != 'subscriber': ... continue ... if data['email'] != 'user(a)example.com': ... continue ... print(data['email']) ... ``` You should not be getting that 'Subscription request already pending' response. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

3 years, 4 months

[MM3-users] Re: Moderation request email from a non existing user

by Mark Sapiro

On 5/31/22 02:15, David Lopez via Mailman-users wrote: > Every day mailman (MM 3.3.3 (Tom Sawyer)) sends me an email with a held unsubcription request from X user. This user is not subcribed to that list anymore but every day an email is sending to the list owner. On the web interface there's nothing pending for approval or similar. Is there anyway to avoid receiving this anoyning email every day? Thanks in advance. Postorius >=1.3.5 has an `Unsubscription Requests` choice in the `Subscription requests` pull down. This choice will list the request with and allow you to delete it. For earlier Postorius this needs to be done via the REST API or via mailman shell. For example: ``` $ bin/mailman shell -l list(a)example.com Welcome to the GNU Mailman shell Use commit() to commit changes. Use abort() to discard changes since the last commit. Exit with ctrl+D does an implicit commit() but exit() does not. The variable 'm' is the list(a)example.com mailing list >>> pendingsdb = getUtility(IPendings) >>> for token, data in pendingsdb.find(m, pend_type='unsubscription'): ... if data['token_owner'] == 'moderator': ... pendingsdb.confirm(token) ... data about the request will be displayed here >>> commit() >>> ``` -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

3 years, 2 months

[MM3-users] Re: Performance issues with Mailman 3

by Mark Sapiro

On 10/6/22 15:08, Wong, Wesley wrote: > Hi, I am having some performance problems with Mailman 3 (version 3.3.5) when it sends out a lot of messages to people. We recently migrated from our old Mailman 2.1.x system to Mailman 3.x about 3 months ago. We host about 350 lists with membership counts varying between 10 to 1500 members on some of the lists. I noticed that outbound messages from Mailman to our outbound SMTP are getting delayed up to an hour when Mailman is processing a lot of messages going to lists. I suspect the difference is your Mailman 3 instance has full VERP enabled and your Mailman 2.1 instance did not. I.e., In MM 2.1, you did not override the default `VERP_DELIVERY_INTERVAL = 0` in mm_cfg.py, but in MM 3 you have `verp_delivery_interval: 1` in the `mta` section of mailman.cfg. > I was wondering if there are good performance tuning tips to be considered for Mailman 3 (i.e. OS, database and/os Mailman Core). I am running the setup on a VM. Here is my configuration for VM and OS: Search the Titles for `performance` at https://wiki.list.org/. The articles you find are all written for MM 2.1, but you're really dealing with the MTA and the issues are the same for MM 3. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

2 years, 9 months

[MM3-users] Re: running a script for the mailman-shell (to discard held messages) gives errors about a package argument for relative import

by Mark Sapiro

On 6/15/24 03:10, via Mailman-users wrote: > > Now, obviously, I put these exact lines in a script, so I don't have to type it out every single time, and try to run it using > >> sudo -u list mailman shell -l testlist.our-server.de -r ./mm_discard_held.py > > but that invariably gives me ... > TypeError: the 'package' argument is required to perform a relative import for './mm_discard_held' To run a script named mm_discard_held.py in that way, the script must define a function mm_discard held. The content needs to be something like ``` from mailman.app.moderator import handle_message from mailman.config import config from mailman.interfaces.action import Action from mailman.interfaces.request import IListRequests # The above are needed. These imports are done automatically when # running mailman shell interactively, but not when running a script. def mm_discard_held(mlist) rdb = IListRequests(mlist) for req in rdb.held_requests: if req.request_type == RequestType.held_message: config.db.commit() ``` See `mailman shell --details` > I've tried moving it to the path where the mailman binary lives, as suggested. That is what you should do. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

1 year, 1 month

[MM3-users] Re: Pck files stuck in out queue

by Mark Sapiro

On 2/7/25 10:08, Lieuallen, Thomas Otis via Mailman-users wrote: > > I have a remaining issue that baffles me. With my test list, I created a template for ‘list:member:regular:header’. It just has text in it, keeping it simple. I create it and see the entries in the database tables and they are correct. However, when I send a test email, my template text is not inserted, but instead another simple text: “testing hello”. I do the same with list:member:regular:footer, and it works just fine. It is just the header that is working incorrectly. When I remove that postorius header template for the list, that ‘testing hello’ goes away. So, it’s only appearing when I have a postorius template setup for that list. What do you get from the following database queries? SELECT * FROM template WHERE name = 'list:member:regular:header'; SELECT * FROM postorius_emailtemplate WHERE name = 'list:member:regular:header'; Depending on your installation, these tables may be in separate databases, i.e., template table in mailman db and postorius_emailtemplate in mailmanweb db. And what do you get if you go to the uri in the result(s) from the first query? The context in the first results and maybe the second results should also be a clue. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

5 months, 3 weeks

[MM3-users] Re: Log rotation

by Mark Sapiro

On 04/10/2018 03:39 PM, Abhilash Raj wrote: > Yes, this is a known thing and the reason for it is that the SIGTERM isn't delivered till the process wakes up next time. We can have the process sleep for a smaller amount of time to fix this, but then it would mean that it will be running much more often. > > It is the Retry runner, which has sleep_time set to 15m. I don't know if you want to retry very often, like 2min or 5min, but you can change it by adding the following snippet to your mailman.cfg A long time ago (in a galaxy far away) Mailman 2.1 had a global DELIVERY_RETRY_WAIT setting. This setting had been ineffective for years, relying instead on the retry runner's sleep time. In 2.1.26, for what turned out to be a probably spurious reason, I made DELIVERY_RETRY_WAIT effective again, although the retry runner still sleeps for 15 minutes. We could implement something similar in MM 3 to allow the retry runner to have a much shorter sleep time without actually retrying the delivery that often. The down side of that is even though we aren't retrying delivery each time the runner wakes up, we are dequeueing and requeueing the message each time and this would have a file system impact. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

7 years, 3 months

[MM3-users] Re: Recommended way to backup a mailman3 instance

by Mark Sapiro

On 4/13/20 3:22 AM, Andrew Hodgson wrote: > > I was working round my own backup strategy a few weeks ago as I discussed earlier on this list. One thing I found when the directory structure that Mailman Core uses is that the lists directories do not get re-created when Mailman Core starts if they don't already exist. Is this something that should happen, and when would those directories get created? Every other directory did get re-created. > > In my case I was moving from a Dockerised installation to one using venvs, and as the directories from the Docker installation were empty as there hadn't been any postings for a few days, I didn't transfer those over. I created the list directories myself and everything carried on. These var/lists/<list-id>/ directories hold the digest.mmdf files which are mailboxes containing the messages for the next digest. A list post will be added to the digest.mmdf file which will be created if necessary, but if the var/lists/<list-id>/ directory doesn't exist, the create will fail. Thus, the var/lists/<list-id>/ directories need to exist. Normally, they are created when the list is created so they need to be part of a backup. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

5 years, 3 months

[MM3-users] Re: Restore/export existing Mailman lists + hyperkitty archives on a new server

by Mark Sapiro

On 9/15/22 23:10, Kimmo L wrote: >> Those errors should result in the affected messages being saved as *.pck >> files in Mailman's var/archives/hyperkitty/spool/ directory. Possibly >> there is only one such 'bad message' which is retried each time there is >> a new post resulting in the same error. How many such files are there? >> If only one, please send me that one too. Also, these can be examined >> with mailman qfile. > There was 3 file. I used the mailman qfile command and the output was this: > [----- start pickle -----] > <----- start object 1 -----> > Traceback (most recent call last): ...> File "/usr/lib64/python3.9/email/generator.py", line 239, in _handle_text > raise TypeError('string payload expected: %s' % type(payload)) > TypeError: string payload expected: <class 'list'> These again are anonymized posts that had important headers removed because of the incorrect anonymous_list_keep_headers setting. > After i cleaned up /var/archives/hyperkitty/spool folder, mailman.log file is much cleaner and i dont see any new no "Message-ID" error at the moment. Yes. Now that you have the default anonymous_list_keep_headers setting things should be OK going forward. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

2 years, 10 months

[MM3-users] Re: Is there a way of recording bounce messages?

by Mark Sapiro

On 9/27/23 06:35, Mohsen Masoudfar wrote: >>> I am interested to know more about the reasons behind these bounces, they are all handled by Postfix and deleted. ...>>> Is there a way to send them to a mailbox? Or prevent them from being deleted, so that I can look into the message body? Presumibly you have settings for Postfix like those at https://docs.mailman3.org/projects/mailman/en/latest/src/mailman/docs/mta.h… or https://docs.mailman3.org/en/latest/install/virtualenv.html#setup-mta The simplest way is to deliver bounces to some user instead of Mailman. This will disable automated bounce processing for the list so it should be done temporarily only to see a few bounce messages and then reverted. Lets say you have a few problem lists, list1(a)example.com and list2(a)example.com. Put the following in Postfix main.cf ``` virtual_alias_domains = example.com virtual_alias_maps = regexp:/etc/postfix/temp-aliases ``` Then create /etc/postfix/temp-aliases with this content ``` list1-bounces(a)example.com user(a)example.net list2-bounces(a)example.com user(a)example.net ``` where user(a)example.net is the address that should receive the bounces. There are also ways to save copies of all mail received by Postfix. See, e.g., https://www.electricmonk.nl/log/2015/03/06/keep-an-archive-of-all-mails-goi… -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

1 year, 10 months

[MM3-users] Re: Weird behavior with Outlook and any MM3 List

by Mark Sapiro

On 11/1/23 09:42, Thomas Ward via Mailman-users wrote: > I'm seeing some *really* odd behavior with Mailman 3 and Microsoft Outlook clients (even Outlook Web Access). > > When messages come in with a subject prefix, of any way shape or form, Microsoft Outlook seems to *completely strip* the prefixes from any "reply" sent, which could be problematic. > > While the "In-Reply-To" header is still intact and such, Microsoft just seems to 'strip' prefixes from subject lines and that causes a problem in some cases. > > Has anyone observed this behavior before? If so, have you found any workarounds? This is a Microsoft/Outlook issue. As you note at https://lists.mailman3.org/archives/list/mailman-users@mailman3.org/message… this is just the way Outlook displays the message subject. I know of now way around this, but I don't use Microsoft MUAs. Does Outlook also strip the prefix from replies? This probably wouldn't matter for replies to the list as the list will add it back, but might be an issue for other direct recipients. Presumably, Outlook is looking at the [] around the prefix. You might try either dropping the brackets or using something else such as _My-List_ instead of [MyList]. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

1 year, 9 months

[MM3-users] Re: Permission denied for fulltext_index

by Mark Sapiro

On 11/30/22 08:41, Jan Eden via Mailman-users wrote: > > Changing the PATH value to a full path (/opt/mailman/fulltext_index) > and restarting mailman-web generates a different error: Yes, it should be a full path. >> ERROR 2022-11-30 16:38:37,953 1028315 django.request Internal Server Error: /archives/list/testing@lists.eden.one/thread/T6MCILAVQYV5QIOGHJB3JNJ26Z7E2Z35/reattach-suggest >> Traceback (most recent call last): >> File "/opt/mailman/venv/lib/python3.10/site-packages/django/core/handlers/exception.py", line 55, in inner >> response = get_response(request) >> File "/opt/mailman/venv/lib/python3.10/site-packages/django/core/handlers/base.py", line 197, in _get_response >> response = wrapped_callback(request, *callback_args, **callback_kwargs) >> File "/opt/mailman/venv/lib/python3.10/site-packages/hyperkitty/lib/view_helpers.py", line 137, in inner >> return func(request, *args, **kwargs) >> File "/opt/mailman/venv/lib/python3.10/site-packages/hyperkitty/views/thread.py", line 454, in reattach_suggest >> sugg_thread = msg.object.thread >> AttributeError: 'NoneType' object has no attribute 'thread' This says that searching the archive for this list for messages matching the current message subject returns None. It is probably a bug that we don't test for this, but what exactly are you doing? See the post at https://lists.mailman3.org/archives/list/mailman-users@mailman3.org/message… for the only way I've been able to create an error like this. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

2 years, 8 months

[MM3-users] Re: Strange URL in error message

by Mark Sapiro

On 7/17/24 11:47 PM, waps(a)tantalus.jena.thur.de wrote: > Hi Mark, > thanks for your patience. But now I get a little spinning wheel on the archive site, 3 "Error 500" in the Browser console and /var/log/mailman3/web/mailman-web.log shows repeatedly the error: > > ERROR 2024-07-18 08:34:38,261 1760 django.request Internal Server Error: /mailman3/hyperkitty/list/latex-team(a)latex-project.org/posted-to > Traceback (most recent call last): ...> File "/usr/lib/python3/dist-packages/django/db/models/query.py", line 435, in get > raise self.model.DoesNotExist( > hyperkitty.models.thread.Thread.DoesNotExist: Thread matching query does not exist. > > Since the archive is still quite empty and we have only one mailing lis, it would not be a big loss if it could resettet. But how I can do the, without resettinh the member data? Could it be that the Sqlite interface is not as stable as one the the "real databases"? As a superuser, you can go to /mailman3/hyperkitty/list/latex-team(a)latex-project.org/ and delete the archive. This will not delete the mailing list, and an empty archive will eventually be recreated, and new posts will be archived. Regarding the issue, see https://gitlab.com/mailman/hyperkitty/-/issues/504. Ensure that cron is running the Django hourly, daily, etc. tasks and that qcluster is running. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

1 year

[MM3-users] Re: How to identify the exact reason behind the shunting of the mails ?

by Mark Sapiro

On 6/18/21 3:34 AM, Shashikanth Komandoor wrote: > Thank you Mark for your immediate response. > > The below are the tracebacks I could find in the mailman log. > > *Jun 17 19:04:08 2021 (8642) Uncaught runner exception: sequence item 0: > expected str instance, Header foundJun 17 19:04:08 2021 (8642) Traceback > (most recent call last): File > "/var/lib/mailman/mailman/src/mailman/core/runner.py", line 173, in > _one_iteration self._process_one_file(msg, msgdata) File > "/var/lib/mailman/mailman/src/mailman/core/runner.py", line 266, in > _process_one_file keepqueued = self._dispose(mlist, msg, msgdata) File > "/var/lib/mailman/mailman/src/mailman/runners/pipeline.py", line 37, in > _dispose process(mlist, msg, msgdata, pipeline) File > "/var/lib/mailman/mailman/src/mailman/core/pipelines.py", line 50, in > process handler.process(mlist, msg, msgdata) File > "/var/lib/mailman/mailman/src/mailman/handlers/avoid_duplicates.py", line > 59, in process addrs = getaddresses(msg.get_all(header, [])) File > "/usr/lib64/python3.6/email/utils.py", line 112, in getaddresses all = > COMMASPACE.join(fieldvalues)TypeError: sequence item 0: expected str > instance, Header foundJun 17 19:04:08 2021 (8642) SHUNTING: > 1623936848.367641+4753f8127953071ddcabfe828ebe72fcf04d0ae2* This is https://gitlab.com/mailman/mailman/-/issues/881 which is fixed by https://gitlab.com/mailman/mailman/-/merge_requests/841 which will be in Mailman 3.3.5 when it's released. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

4 years, 1 month

[MM3-users] Re: There are currently no mailing lists!?

by Mark Sapiro

On 7/22/24 11:05, Christian via Mailman-users wrote: > I recently upgraded Debian 10 'buster' and when I logged in to Postorius in > the browser I got the ' Mailman REST API not available. Please start Mailman > core' message. So I started it up from the command line doing 'sudo -u > mailman /opt/mailman/mm/venv/bin/mailman start -f' which yielded > > Starting Mailman's master runner > Generating MTA alias maps > mailman@lists: > > Now when I do ' https://lists.ccalternatives.org/mailman3/lists/?all-lists' > in my browser I get the message ' There are currently no mailing lists.' > Meanwhile their archives exist. Are the lists actually gone or ? I suspect the lists are not gone. I suspect mailman core is not using the correct database. I also note you started mailman core with sudo -u mailman /opt/mailman/mm/venv/bin/mailman start -f Are you using the Debian packages or did you install Mailman following https://docs.mailman3.org/en/latest/install/virtualenv.html Assuming the latter, what do the following show? sudo -u mailman /opt/mailman/mm/venv/bin/mailman lists sudo -u mailman /opt/mailman/mm/venv/bin/mailman info In particular does `mailman info` report the correct path to mailman.cfg and the appropriate database? -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

1 year

[MM3-users] Re: bounce_score_threshold control

by Mark Sapiro

On 2/24/23 08:08, Tibor Molnar wrote: > > I have recently started receiving these mails from the system. First it was frighting me out that people's account got disabled, but actually nothing has happened, nobody got disabled. I however looking to figure out how can I control or change settings for this particular mechanism. Please advise me. > > emailname(a)example.it's subscription has been disabled on listname(a)domain.eu due to their bounce score > exceeding the mailing list's bounce_score_threshold. I'm not sure what you are looking at, but when you receive that message, it means the user's Delivery status has been set to Disabled by Bounces and the user will be sent `Bounce disable warnings` (default 3) warnings at intervals of `Bounce disabled warnings interval` (default 7 days) and if nothing is done the user will then be removed from the list. The notice you received, at least since Mailman 3.3.5, should contain a copy of the bounce DSN so you can determine if this user's address is actually undeliverable. Settings that control this are in the Postorius admin UI under Bounce Processing. The only other control is in mailman.cfg in the [mta] section where you can set `verp_probes: yes` See the docs at https://docs.mailman3.org/projects/mailman/en/latest/build/lib/mailman/mode… -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

2 years, 5 months

[MM3-users] Re: Mailman 3 on Ubuntu 18.04

by Mark Sapiro

On 12/9/18 12:53 PM, Odhiambo Washington wrote: > On Sun, 9 Dec 2018 at 22:37, Mark Sapiro <mark(a)msapiro.net> wrote: >> >> The nginx logs generally aren't too useful. Both Mailman core and Django >> write logs. Core's log is probably in var/log/mailman.log where is >> defined in mailman.cfg as var_dir. Django's log is defined in Django's >> settings in the LOGGING setting. >> > > Which file contains Django's settings? See <http://docs.list.org/en/latest/pre-installation-guide.html#django-knowledge>. >>> Ultimately I need to figure out: >>> 1. How the mailman queue runners are started and stopped >> >> mailman start >> > > Why not systemctl mailman start?? Just wondering, as that is what I am > seeing is common in Ubuntu. > Does this make the components start after a reboot? 'mailman start' is the command that starts Mailman. When you install Mailman, you typically create an init, upstart or systemd script to enable 'systemctl mailman start'. Presumably the Ubuntu mailman3-full package installs such scripts for Mailman and other components that need to be started at boot, but I'm not that familiar with this package. >>> 2. Migrate s few lists from Mailman2 to this Mailman3 >> >> mailman import21 >> <django management> hyperkitty_import >> > > Please show me where that is documented. I have to pull my lists from an > old machine to this new one. Those commands have --help options. There is also some documentation at <https://mailman.readthedocs.io/en/latest/src/mailman/commands/docs/import.h…> and <https://hyperkitty.readthedocs.io/en/latest/install.html#importing-the-curr…>. All of our formal documentation is at <http://docs.list.org/en/latest/>. > So, when you talk about Django settings, you are referring to mailman.cfg > file??? :-) No. I am talking about a file often named settings.py and which my in turn import a file named something like settings_local.py. See my first answer above. > I have spent all my weekend till now, trying to figure all this out. Life > seems a little easier with mm2.x .... I expect some of that is just what you're familiar with, but yes, Mailman 3 is more complex than Mailman 2.1 just as Mailman 2.1 is more complex than a Bcc: list. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

6 years, 7 months

[MM3-users] Re: "Delete this thread" -> 'HTTP-500 -> NoneType' object has no attribute 'all'

by Mark Sapiro

On 05/28/2017 12:40 PM, Simon Liebold via Mailman-users wrote: > I didn't find a 'delete message' button. But "Delete this thread" went > straight to the 500. Both, Hyperkitty and django-mailman are up-to-date > with master. 'Delete this message' is in the 'message' view. "Delete this thread" is in the 'thread' view. > ERROR 2017-05-28 13:04:14,496 29835 django.request Internal Server > Error: > /hyperkitty/list/simsimmer@lists.simonliebold.de/thread/4I2TRZTFXMMH2XRQAS73IAEIYROFE5YM/delete > Traceback (most recent call last): > File > "/opt/mailman/venv-2.7/local/lib/python2.7/site-packages/Django-1.11rc1-py2.7.egg/django/core/handlers/exception.py", HyperKitty requires Django<1.11. The comment in the requirements in setup.py implies this is only a django-haystack requirement "Django>=1.8", # Don't upgrade to 1.11 until django-haystack supports it: # https://github.com/django-haystack/django-haystack/pull/1499 "Django<1.11", but since it works for me with Django 1.10, you might want to try that. ... > File > "/opt/mailman/venv-2.7/local/lib/python2.7/site-packages/Django-1.11rc1-py2.7.egg/django/forms/models.py", > line 1168, in __deepcopy__ > result.queryset = self.queryset.all() > AttributeError: 'NoneType' object has no attribute 'all' In particular, this looks like Django 1.11 bug <https://code.djangoproject.com/ticket/27975>. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

8 years, 2 months

[MM3-users] Re: Change digest frequency and default threshold

by Mark Sapiro

On 10/11/2018 03:48 AM, Henrik Rasmussen wrote: > > Before running /usr/bin/docker exec mailman-core mailman digests --periodic --verbose > > Every 2.0s: ls -l;grep "Test 20181011" digest.mmdf;tail /var/log/mailman/core/smtp.log Thu Oct 11 12:23:32 2018 > > -rw-rw---- 1 100 nogroup 103433 Oct 10 14:36 digest.1.7.mmdf > -rw-rw---- 1 100 nogroup 155540 Oct 11 11:29 digest.1.8.mmdf > -rw-rw---- 1 100 nogroup 104002 Oct 11 12:21 digest.mmdf ... > After running /usr/bin/docker exec mailman-core mailman digests --periodic --verbose > > Every 2.0s: ls -l;grep "Test 20181011" digest.mmdf;tail /var/log/mailman/core/smtp.log Thu Oct 11 12:44:45 2018 > > total 372 > -rw-rw---- 1 100 nogroup 103433 Oct 10 14:36 digest.1.7.mmdf > -rw-rw---- 1 100 nogroup 155540 Oct 11 11:29 digest.1.8.mmdf > -rw-rw---- 1 100 nogroup 104002 Oct 11 12:21 digest.1.9.mmdf The mailman digests --periodic command renames the digest.mmdf file to (in this case) digest.1.9.mmdf and then queues an entry in the digest queue for DigestRunner to actually build the digests from the digest.1.9.mmdf file, send them and remove the digest.1.9.mmdf file. Something is going wrong in this process in DigestRunner. Is DigestRunner running? Is there anything in Mailman's digest or shunt queues? Is there anything logged in Mailman's 'mailman' log? -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

6 years, 9 months

[MM3-users] Re: Importing users and mas settings "Moderation Action"

by Mark Sapiro

On 5/19/22 06:17, bob B via Mailman-users wrote: > I imported email lists and users to version 3 from version 2. One of my lists all the users are set to 'Hold for moderation" for "Moderation Action". I assume this is because they were set to moderation on mailman 2? Correct. > But if I mass subscribe users via the Gui in mailman 3 then their default "Moderation Action" is set to "none" so I assume it follows list default, is that a correct assumption? Yes. > I believe there is no way to import mass users and set their "Moderation Action" at the same time is this correct? Wasn't that possible in mailman 2? That's correct, and that was also the case in Mailman 2.1 . > is there a way to change all their user's "Moderation Action" for a list in bulk? You could do it with `mailman shell`, but not in Postorius. > So it looks like for any users I mass subscribe, "Moderation Action" will be set to "none" and if I want to change that I have to do it manually for each user? If you are a list owner whose only access is via Postorius, that's correct. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

3 years, 2 months

[MM3-users] Re: Welcome message template errors

by Mark Sapiro

On 4/27/24 18:44, Mark wrote: > > I created a text file as a custom welcome message and chowned it to > mailman:mailman, then edited the database record (template.uri) from > > http://localhost:8000/mailman3/api/templates/list/mytestlist.mydomain.com/l… > TO > file:///home/mark/welcome.txt > > This delivered the custom welcome message to the user when subscribed > via Mass Subscribe. > > I'm not sure what this tells me other than a call can be made to a file > but not the database. The retrieval of a file does not involve calling requests.get() to retrieve the URL via http which is where the issue occurs. There's another way to set templates in the file system. If there is no entry in the template table matching the requested template name and context the file system is searched according to the algorithm described at <https://gitlab.com/mailman/mailman/-/blob/master/src/mailman/utilities/i18n…>. $template_dir in that description is Mailman's var/templates/ directory. So for example, to create a lists specific english language welcome template for the list.example.com list you could just put the template in var/templates/lists/list.example.com/en/list:user:notice:welcome.txt or to make it sitewide, in var/templates/site/en/list:user:notice:welcome.txt -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

1 year, 3 months

[MM3-users] Re: Pck files stuck in out queue

by Mark Sapiro

On 2/10/25 12:09, Lieuallen, Thomas Otis via Mailman-users wrote: > > I haven’t figured out how to work with these cache files. I found some mention that a runner would clean up cache files after a week (default) and that one could use ‘cache_life’ to change that default. So, I set the following and restarted mailman. The cache files hadn’t been removed after > 4 hours, so I unset it and restarted again. > > Cache_life: 1d The setting is actually `cache_lifetime` in the `[mailman]` section of mailman.cfg, however when a file is cached, it's lifetime is saved in the database and won't be changed by changing that setting. > Without further information or direction, I’m tempted to just remove those cache files and move on with life. Removing them is OK, but a database entry will still exist pointing to them and attempt to retrieve them will return `Cache content lost`. The way to properly remove them is in Mailman shell ``` $ mailman shell Welcome to the GNU Mailman shell Use commit() to commit changes. Use abort() to discard changes since the last commit. Exit with ctrl+D does an implicit commit() but exit() does not. >>> cm = getUtility(ICacheManager) >>> cm.clear() >>> commit() >>> ``` -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

5 months, 3 weeks

[MM3-users] Re: Link another address

by Mark Sapiro

On 5/18/20 5:13 PM, Brian Carpenter wrote: > I am thinking about incorporating the following Postorius/Django account feature into Affinity: > > Other emails: (no other email) Link another address > > However I cannot think of a serious real-world need for this. What are the developers' reasoning behind this? Was this feature highly demanded by Mailman 3 users? Just trying to find out what I am missing here. In mailman 2.1 there was no concept of user. There were only lists and list members and list members were just email addresses that had no connection to other list members even if they happened to have the same email address. In mailman 3 we have users. Users can have multiple addresses. a user or a particular user's address can be a regular member and an owner of one list, a moderator of another list and a non-member of a third list. Some other address of that user can be a digest member of the first list. All of these addresses and roles can belong to a single user and be managed from that user's one account. Sure, each address can be separate with it's own login, but isn't it more convenient if they are all linked under one login if they all belong to one individual. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

5 years, 2 months

[MM3-users] Re: Hyperkitty "All Threads" results in error

by Mark Sapiro

On 10/14/20 11:12 AM, Mark Dadgar wrote: >> On Oct 4, 2020, at 10:34 AM, Mark Dadgar <mark(a)pdc-racing.net> wrote: >> >> So I have been wrestling with this problem for a while now. Most times, but not every time, when someone clicks on the All Threads link in hyperkitty, a 500 error is generated. Does this happen on all lists, only some lists or only one list? >> I googled it, obviously, and at first I thought it was related to this issue, but that has not proven to be the case: >> >> https://lists.mailman3.org/archives/list/mailman-users@mailman3.org/thread/… What did you google? This query gets a lot of stuff with some hints, but nothing definitive. django render raises VariableDoesNotExist("Failed lookup for key " One thing I noted at <https://askbot.org/en/question/3095/django-error-variabledoesnotexist-faile…> was models out of date - run python manage.py syncdb and python manage.py migrate You could try that. Note that your django admin command might be something other than `python manage.py`. Also note syncdb has been removed from Django >=1.9. Currently, the above becomes `makemigrations` followed by `migrate`. I also suggest running the django admin commands `collectstatic --clear --noinput` and `compress` -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

4 years, 9 months

[MM3-users] Re: Mysterious Bounces Being Logged

by Mark Sapiro

On 3/26/21 11:31 AM, Brian Carpenter wrote: > I keep seeing bounces being recorded in logs/bounce.log that have no > corresponding entry in Postfix's mail.log. For instance one address had > registered about 2 dozen bounces in the past week in bounce.log and yet > every single entry for this address in mail.log has a status=sent. What > am I missing here? They are being bounced by a downstream MTA. If you want to know why, you have a few choices. You can mail the address yourself and hopefully get a DSN back. You can install this patch: ``` > diff --git a/src/mailman/runners/lmtp.py b/src/mailman/runners/lmtp.py > index 6a0bdcf5e..eca474089 100644 > --- a/src/mailman/runners/lmtp.py > +++ b/src/mailman/runners/lmtp.py > @@ -217,6 +217,8 @@ class LMTPHandler: > slog.debug('%s subaddress: %s, queue: %s', > message_id, canonical_subaddress, queue) > status.append('250 Ok') > + if queue == 'bounces': > + config.switchboards['bad'].enqueue(msg, msgdata) > except Exception: > slog.exception('Queue detection: %s', msg['message-id']) > config.db.abort() ``` which will add all received bounces as queue files in Mailman's var/queue/bad. At some point, we need to implement the attachment of the bounce DSN to the notices sent to the admin, but that isn't done yet. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

4 years, 4 months

[MM3-users] Re: Pck files stuck in out queue

by Mark Sapiro

On 2/6/25 12:42, Lieuallen, Thomas Otis via Mailman-users wrote: > Running Mailman 3.3.10 > > I have 11x pck files stuck in my out queue. There tends to always be a .bak file as well. If I run ‘ls -l’, the files are continually updated and rotated through, just as fast as I can look. Are old ones being delivered and new ones added or the same ones just being renamed? I suspect the latter and I suspect the issue in that case is probably due to the OUT runner trying to access a template file like list:member:{regular|generic}:{header|footer}.txt and not having permission. See the last two posts in the thread at https://lists.mailman3.org/archives/list/mailman-users@mailman3.org/thread/… > I believe they’re all for one list. When I use ‘strings’ on one to get a hint of what’s inside, I see the list members at the bottom (recipients). One of them consistently has ‘!’ before their email address. Not sure if that’s a clue. Just no other reason to suspect that, though. The files are python pickles and `strings` shows lots of stuff. The ! may be something in the pickle that's not part of the email address. The way to examine these files is with the `mailman qfile` command. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

5 months, 3 weeks

[MM3-users] Re: footer template priority

by Mark Sapiro

On 6/19/25 12:06, Sam Darwin via Mailman-users wrote: > It seems templates can be configured in the filesystem or via postorius. When I place a footer template in the filesystem it worked fine. Then without adjusting any configuration, without removing that, I tried to use the postorius control panel to add a footer template, which ends up in the database, both the core and web databases. > > But emails stop getting sent out. Emails were posted into Hyperkitty, but not mailed out via postfix. This is a known issue when header and/or footer templates are not accessible. > I checked on the server for "queued emails" and did not find those either, in postfix or mailman. Mysterious. The phenomenon is repeatable on this server. Add a footer template, emails stop. The mails should be in Mailman's var/queue/out queue. Since the issue only occurs in your case when you have a Postorius template, I suspect the issue is in your setting for POSTORIUS_TEMPLATE_BASE_URL. When you have the Postorius template set, what does the database query ``` SELECT uri FROM template WHERE name = 'list:member:regular:footer' and context = 'your.list.id'; ``` return for the uri? is it retrievable? The initial part of the uri up to but not including `/mailman3/api/...` comes from the POSTORIUS_TEMPLATE_BASE_URL setting -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

1 month, 1 week

[MM3-users] Re: Walk-thru for Ubuntu Bionic Mailman 3 Full setup

by Mark Sapiro

On 4/11/19 12:58 PM, brian(a)emwd.com wrote: > > I have this in the apache2.conf file: > > ServerName localhost > > I can access the apache default page when using the following: > > curl http://localhost.com OK, so you can get to Apache, but can you proxy to uwsgi? Try curl http://localhost/mailman3/hyperkitty/api/mailman/archive or curl http://mailman3-lists.com/mailman3/hyperkitty/api/mailman/archive >> Set base_url in mailman-hyperkitty.cfg to use an acceptable host name >> instead of localhost or adjust your apache config to accept localhost. > > Isn't ServerName enough? It should be. The error is apparently elsewhere. > This error also appears in the apache log file when new posts are sent to the list: > > [Thu Apr 11 19:55:38.755750 2019] [:error] [pid 21370:tid 139988922513152] (32)Broken pipe: [client ::1:49574] sending data to httpd-UDS:0 failed The broken pipe is apparently from the exception in mailman-hyperkitty and the 'sending data to httpd-UDS:0 failed' is I think a normal response from uwsgi under this circumstance. I don't know what the issue is, but it is something in the Apache/uwsgi configuration. I note that the client in the above error is an IPv6 loopback address. That may be significant. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

6 years, 3 months

[MM3-users] Re: 'Failed adding message' error when migrating a list from mailman2 to mailman3

by Mark Sapiro

On 10/14/21 9:04 AM, Stephen J. Turnbull wrote: > Prasanth Nair writes: > > > *2%Failed adding message <mailman.2.1593699619.32248>: 'ascii' codec can't > > encode character '\ufffd' in position 252: ordinal not in range(128).* ... > > I ran the check_hk_import script before migrating and it was not > > giving any error. > > In order to catch this kind of error, the check script would have to > pretty much do the import but throw it away. The check_hk_import script <https://www.msapiro.net/scripts/check_hk_import> is designed to detect and report conditions which would cause the django hyperkitty_import process to abort and stop processing the mbox. It doesn't catch errors like the above, but these are non-fatal and just cause the offending message to be skipped. As Steve says, the Message-ID was reported so find it in the mbox and if you want it in the archive, extract it from the mbox, fix it and then just rerun the import with just that message as input[1] and a --since=1970-01-01 option. Note the issue could be in a header, but also could be in a body part implicitly or explicitly declared as charset="us-ascii". -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

3 years, 9 months

[MM3-users] Re: question about recommended postfix configuration

by Mark Sapiro

On 10/02/2017 05:37 AM, Marc Sluiter wrote: > > I got almost everything working besides confirm mails (see my other mail), but I had to use another postfix configuration than recommended, I’m wondering why the recommended way didn’t work. Sidenote: I’m not a postfix expert. Neither am I ;) > This is recommended: > > transport_maps = > regexp:/opt/mailman/core/var/data/postfix_lmtp > local_recipient_maps = > regexp:/opt/mailman/core/var/data/postfix_lmtp > relay_domains = > regexp:/opt/mailman/core/var/data/postfix_domains > > This didn’t work, all mails to mailman were rejected. Postfix did not even check the “local_recipient_maps”. After a lot of reading that made sense to me, because when I understand it correctly, “local_recipient_maps” is only checked for domains in “mydestinations”. And there I only have localhost. On mail.python.org which is a complex Postfix environment with both Mailman 2 and Mailman 3 lists in the @python.org domain, for Mailman 3 I have just transport_maps and relay_domains as recommended (but as hash tables) I don't have anything for Mailman in local_recipient_maps, but python.org is in virtual_alias_domains (not in virtual_mailbox_domains). I think it could be either, and virtual_alias_maps (or virtual_mailbox_maps) does not require anything for Mailman because transport_maps handles it. I think the bottom line here is the list domain must be in one of mydestination or virtual_*_domains. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

7 years, 9 months

[MM3-users] Re: including a template failed

by Mark Sapiro

On 9/23/20 3:03 PM, Wolfgang Bock via Mailman-users wrote: > > Dear Mark, > > Q: Can you get > https://webmail.domain-02.de/mailman3/postorius/api/templates/list/mylist.li > sts.domain-03.de/list:member:regular:footer on the Mailman machine? > > A: No > > Q: What is your setting for POSTORIUS_TEMPLATE_BASE_URL? > > A: > was: POSTORIUS_TEMPLATE_BASE_URL = 'http://localhost/mailman3/' > > Changed it to: POSTORIUS_TEMPLATE_BASE_URL = 'https://localhost/mailman3/' Can you get https://localhost/mailman3/postorius/api/templates/list/mylist.lists.domain… on the Mailman machine? > Now another problem occurs: > > Sep 23 22:51:34 2020 (13492) Cannot connect to SMTP server localhost on port > 25 > HTTPSConnectionPool ... 'localhost' doesn't match either of > 'blog.domain-04.de', 'lists.domain-04.de', 'mail.domain-04.de', > 'domain-04.de', 'www.domain-04.de'"))) The 'HTTPSConnectionPool' message has nothing to do with the 'Cannot connect' issue. > Port 25 entry results from the manual regarding the postfix configuration: > > https://docs.mailman3.org/projects/mailman/en/latest/src/mailman/docs/mta.ht > ml > > e.g.: > [mta] > incoming: mailman.mta.postfix.LMTP > outgoing: mailman.mta.deliver.deliver > lmtp_host: mail.example.com > lmtp_port: 8024 > smtp_host: mail.example.com > smtp_port: 25 And I wonder why it's trying to connect to localhost when smtp_host is mail.example.com. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

4 years, 10 months

[MM3-users] Re: {Spam?} Server name in new account confirmation email

by Mark Sapiro

On 12/3/20 9:26 AM, Meehan, John wrote: > Hi, when someone creates a new account (Sign Up), an email is sent to them to confirm their email address. It has the form below. Does anyone know how I can change the server name in the URL (From server.example.com to something else)? I have an SSL proxy server that I want to direct users through. Thanks. - John This message comes from django_allauth. The domains come from the django_site database table. If you have Postorius >= 1.3.3 The Domains view will show you the SITE_ID for the domain and has an `Edit` link the the Django admin UI entry for that site. Otherwise, you can go to the Django admin `Sites` view and see the DOMAIN NAME and DISPLAY NAME for each defined site, and you can edit the domain name. You may also need to adjust the SITE_ID setting in your Django settings to match the site you want. The trick here is the SITE_IDs aren't displayed in the Django admin UI or in Postorius < 1.3.3, and they may not be sequential or start at 1. If you query the database with `SELECT * FROM django_site;` you'll see entries with columns id, domain and name which will tell you the SITE_ID. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

4 years, 7 months

[MM3-users] Re: mm3 floods syslog

by Mark Sapiro

On 10/31/21 4:18 AM, Wolfgang Bock via Mailman-users wrote: > Dear friends, > > I got another problem: mailman floods my syslog. > > I deleted a domain/webdomain (further mentioned as deleted.org) in > postorius, django web. No entries are now in > /var/lib/mailman3/data/postfix_domains or postfix_lmtp or > /var/lib/mailman3/lists. > > > Here an extract of my syslog (anonymized): > Oct 31 10:49:02 myserver mailman3[1284]: Oct 31 10:49:02 2021 (1284) > Certificate did not match expected hostname: localhost. Certificate: > {'subject': ((('commonName', 'deleteted.org'),),), 'issuer': > ((('countryName', 'US'),), (('or > ganizationName', "Let's Encrypt"),), (('commonName', 'R3'),)), 'version': 3, > 'serialNumber': '0326A733CA7B16D13C830EE707EA614E25F9', 'notBefore': 'Oct 30 > 20:21:08 2021 GMT', 'notAfter': 'Jan 28 20:21:07 2022 GMT', > 'subjectAltName': (('DN > S', 'deleteted.org'), ('DNS', 'lists.deleteted.org'), ('DNS', > 'mail.deleteted.org'), ('DNS', 'ns.deleteted.org'), ('DNS', > 'www.deleteted.org')), 'OCSP': ('http://r3.o.lencr.org',), 'caIssuers': > ('http:/ > /r3.i.lencr.org/',)} > > This appears nonstop. Your ssl certificate is for the domain `deleted.org` with alt names `lists.deleted.org`, `ns.deleted.org`, `mail.deleted.org` and `www.deleted.org`. Django is accessing things at a URL for a different domain. Get a new certificate for the domain(s) you are actually using. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

3 years, 9 months

[MM3-users] Re: Mailman3 doesn't send Mails to users (Debian12 with Plesk)

by Mark Sapiro

On 5/27/24 1:55 AM, marcus--- via Mailman-users wrote: > > What I also notice is that it is not yet storing anything in the archives. Could this be a setting that has changed, or an incorrect authorization (/var/lib/mailman3/archives/ belongs to list but is empty) or is the archive creation only running via a cronjob that perhaps still needs to run? There are two things stored in that directory If the prototype archiver is enabled, each list which has prototype checked as an Active archiver will have messages stored in maildir format in archives/prototype. If the hyperkitty archiver is configured and enabled, when mailman-hyperkitty fails to send a message to hyperkitty, it queues that message in archivers/hyperkitty/spool. If there no current failed messages, that directory will be empty, and if there have never been failures, that directory may not exist. Enabling archivers is done in mailman.cfg similarly to ``` [archiver.hyperkitty] class: mailman_hyperkitty.Archiver enable: yes configuration: /opt/mailman/mm/mailman-hyperkitty.cfg [archiver.prototype] enable: yes ``` > I don't need Usenet, thanks :-) Then no list should have it's Gateway to mail or Gateway to news setting = Yes. Also, you can disable nntp runner by putting ``` [runner.nntp] start: no ``` in mailman.cfg -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

1 year, 2 months

[MM3-users] Re: Mailman3 breaking DKIM signatures

by Mark Sapiro

On 1/21/25 4:40 AM, jefdriesen(a)hotmail.com wrote: > Hi, > > I have configured a mailman3 mailinglist to not modify the headers and body (e.g. no footer, no subject prefix, etc) to avoid breaking DKIM signatures. This worked well, until I encountered an email where the DKIM signature was still broken. After some debugging, the problem is caused by an attachment, where the mime header is changed by mailman3. ... > The Content-Type header is modified (word-wrapped), and this breaks the DKIM signature. > > How can I prevent this from happening and configure mailman to not modify the email? You could try a patch like ``` --- a/src/mailman/runners/lmtp.py +++ b/src/mailman/runners/lmtp.py @@ -186,7 +186,8 @@ class LMTPHandler: listnames = set(getUtility(IListManager).names) # Parse the message data. If there are any defects in the # message, reject it right away; it's probably spam. - msg = email.message_from_bytes(envelope.content, Message) + msg = email.message_from_bytes(envelope.content, Message, + policy=email.policy.Compat32(max_line_length=0)) msg.set_unixfrom(envelope.mail_from) except Exception: elog.exception('LMTP message parsing') ``` This should prevent folding headers, but this is problematic. It may be that the incoming message contains a folded header and this change will not fold it in the outgoing message. I think not, but I don't know for sure. You'll have to test this. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

6 months, 1 week

[MM3-users] Re: Mailing lists sometimes do not honor moderation_action of non-members

by Mark Sapiro

On 10/31/19 2:21 AM, Alan So wrote: > We notice that sometimes mailing lists do not honor moderation_action of non-members. Instead, it follows the default action of the lists. It looks like we can sometimes fix the issue by adding/removing the non-members or adding/removing them as owner in certain order but we are not sure about exactly what we have to do. So, it looks like that it is not caused by the list configuration. We have already verified from the web UI, mailman client and the DB table member that moderation_action of the corresponding non-members are set correctly. Any clues for us to provide further information to troubleshoot the issue and any recommended workaround to get around the issue? Thanks a lot. Reasons why a post From: a non-member would not be treated according to that non-members moderation_action include the following: If any address in any of the places configured in mailman -> sender_headers (default = From: header, envelope sender, Reply-To: header, Sender: header) is that of a member, that member's moderation_action will apply. If any of the addresses above matches an address or pattern in one of the lists *_these_nonmembers attributes that action will apply rather than the corresponding non-member's moderation_action. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

5 years, 9 months

[MM3-users] Re: Example Logrotate files for Mailman3

by Mark Sapiro

On 6/30/20 2:10 PM, Andrew Hodgson wrote: > Hi, > > Does anyone have any usable Logrotate files for Mailman3 (especially for the web stuff)? Here are two. They differ in the name and location of the Django log. Both of them signal gunicorn to reopen its logs. If you are using uWSGI or some other WSGI support, you'll need to figure that out. In this one, all the logs including Django's are in /opt/mailman/mm/var/logs/ > /opt/mailman/mm/var/logs/*.log { > missingok > sharedscripts > su mailman mailman > postrotate > /opt/mailman/mm/bin/mailman reopen &>/dev/null || true > if [ -r /opt/mailman/mm/var/gunicorn.pid ]; then > PID=`cat /opt/mailman/mm/var/gunicorn.pid` > kill -s USR1 $PID > fi > endscript > } This one is actually two because the Django log is in /var/log/ rather than /opt/mailman/mm/var/logs/. > /opt/mailman/mm/var/logs/*.log { > missingok > sharedscripts > su mailman mailman > postrotate > /opt/mailman/mm/bin/mailman reopen &>/dev/null || true > if [ -r /opt/mailman/mm/var/gunicorn.pid ]; then > PID=`cat /opt/mailman/mm/var/gunicorn.pid` > kill -s USR1 $PID > fi > endscript > } > } > /var/log/mailman-web/*.log { > missingok > notifempty > delaycompress > su mailman mailman > } -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

5 years, 1 month

[MM3-users] Re: Send MySQL data to mailman list for distribution

by Mark Sapiro

On 6/1/21 1:46 PM, Zden Pos via Mailman-users wrote: > I'm trying to connect existing MySQL data to the Mailman distribution list. > Actually, a message is sent from the web page. It remains on this page and is also stored in the MySQL table on the provider server. This is working fine. > I have also existing Mailman distribution list generated from the cPanel on the provider server as well. This list is working fine when used for a direct mailing. > > I need to link the MySQL table to the Mailman distribution list. When the message is posted on the web page and stored in the MySQL table, it is also to be sent to members as per the Mailman distribution list. I don't know where to configure it. In the phpMyAdmn page of the MySQL or in the Malman menu ? I am very confused because it appears that your Mailman list with MySQL data is a Mailman 3 list, but cPanel lists as far as I know are Mailman 2.1 only. My best guess is you want posts to the Mailman 3 list to be forwarded to thwe cPanel list. If so, just add thee cPanel list's posting address as a member of the Mailman 3 list. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

4 years, 2 months

[MM3-users] Re: Moving from forum to Mailman 3

by Mark Sapiro

On 10/9/20 4:39 PM, jonathan.mailing.lists(a)gmail.com wrote: > > Thanks for this. I’m sure I could turn that basic information (from, date, subject) into an mbox file somehow. What else does Mailman need, eg to work out which messages are in reply to which? To properly thread messages, each message should have a unique Message-ID: header and all replies should have an In-Reply-To: header with the Massage-ID: of the message to which it's a reply. E.g. the first message in a thread has Message-ID: <some_unique_string(a)example.com> A reply to that has Message-ID: <different_unique_string(a)example.com> In-Reply-To: <some_unique_string(a)example.com> and a reply to that message has Message-ID: <third_unique_string(a)example.com> In-Reply-To: <different_unique_string(a)example.com> The angle brackets around the strings are required and every Message-ID: in the entire archive for a list must be unique. If threading is not an issue, hyperkitty_import will generate a unique Message-ID: for any message without one. References: headers are not required. > Would it be possible to add ex-forum message “emails” like this to an existing mailing list archive? Yes. When running hyperkitty_import you probably need to specify since= in this case, otherwise only messages with dates newer that the newes message in the existing archive will be imported. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

4 years, 9 months

[MM3-users] Re: Fwd: [Django] ERROR (EXTERNAL IP): Internal Server Error: /mailman3/hyperkitty/api/mailman/archive

by Mark Sapiro

On 3/2/21 12:10 AM, David Partain via Mailman-users wrote: > > I'm not on the latest mailman, unfortunately. I use the bundled packages on Ubuntu 18.04, which are > > mailman3 3.1.1-9 > mailman3-doc 3.1.1-9 > mailman3-full 3.1.1-9 > mailman3-web 0+20170523-14 > >>From the mailman-web logs, it looks like it's using python2. That would explain a lot. In particular, it would explain why the .pyc files aren't located per PEP 3147 <https://www.python.org/dev/peps/pep-3147/>. But it also raise the question of how anything works at all. Python >= 3.6 is required. HyperKitty and Postorius used to run under Python 2, but haven't for some time. I'm not sure what versions are in Debian/Ubuntu 0+20170523-14 package. > Put the symlink back, and it works again. (But still without the emojis working, of course.) > > Postgres is what's in settings.py, but I assume that should be overriden by what's in settings_local.py. > > Baffling. I don't really know what to do next, but I'll keep looking. I suggest removing all the .pyc files and restarting the services. That will probably recreate .pyc files, but they should then be good. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

4 years, 4 months

[MM3-users] Re: Subscribe to a list on a Mailman3 server without account registration

by Mark Sapiro

On 6/6/23 01:07, c.buhtz(a)posteo.jp wrote: > Hello, > I might miss something here but I'm confused about a difference between Mailman2 and 3. > > On MM2 I just subscribed a mailinglist via my EMail address. I didn't need to create an account on the MM2 server first. > But it seems on MM3 it is otherwise. For example I use MM3 on python.org. It seems I have to register first before subscribing to a mailing list. > https://mail.python.org/mailman3/lists/twisted-web.python.org/ You can subscribe without creating an account. You do need an account if you want to manage your subscription via the web UI or to visit private archives, but not to just be a list member or visit public archives. The issue with the https://mail.python.org/mailman3/lists/ lists is that the anonymous subscribe form which looks like: ----------------------------------------- You can also subscribe without creating an account. If you wish to do so, please use the form below. Your email address __________ Your name (optional) __________ ----------------------------------------- has been removed from list info page because it was being abused by people attempting to subscribe third parties. You can still subscribe without creating an account by emailing twisted-web-subscribe(a)python.org. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

2 years, 1 month

[MM3-users] Re: Problems with signed messages from Apple Mail

by Mark Sapiro

On 11/12/21 6:06 AM, Rainer Herbst wrote: > Dear all, > > after upgrading his Mac to MacOS 12.0.1 Monterey, one of our users can > not send send signed messages to lists using Apple Mail v. 15.0 > (3693.20.0.1.32). > > mailman3 denys the delivery of the mail with error message: > host 172.19.0.4[172.19.0.4] said: 501 Message has defects > (in reply to end of DATA command) > > I.e. mailman does not process the message, but rejects it already in the > incoming smtp session. Is that a known issue? Incoming SMTP is relaying the message via LMTP to Mailman's LMTP runner which in turn parses the message using Python's email.message_from_bytes() method and the resultant message has defects. Rejecting such a message outright as we do may be too harsh, but it is hard to say without seeing the message and what Python's email package thinks the defects are. This could be a bug in Apple Mail or in the Python email module. If you wish to see the details, you could do the following in `mailman shell` ``` import email from mailman.email.message import Message content = b"""\ (paste the raw message here) """ msg = email.message_from_bytes(content, Message) msg.defects ``` This will show you what the defects are. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

3 years, 8 months

[MM3-users] Re: mailman3-suite not sending confirmation emails

by Mark Sapiro

On 1/2/19 11:59 PM, ccehb(a)nus.edu.sg wrote: > Hi, > > This is my first experience using mailman. I recently downloaded and setup mailman3 suite which comes with hyperkitty. I don't really understand why the email is not working. I had an external smtp server (mymailserver.com) residing on different IP from my web server that hosts mailman stuff. > > I had modified settings.py with debug = False and added the following: > > EMAIL_BACKEND = 'django.core.mail.backends.smtp.EmailBackend' > EMAIL_HOST = 'mymailserver.com' > EMAIL_PORT = 25 These should be OK, but they are only for messages from Django, not for list mail. Also if mymailserver.com requires authentication, you may need to set EMAIL_HOST_USER and EMAIL_HOST_PASSWORD. > I also added these inside mailman.cfg > > [mta] > outgoing: mailman.mta.deliver.deliver > lmtp_host: 127.0.0.1 > lmtp_port: 8024 The above are all defaults. > smtp_host: mymailserver.com > smtp_port: 25 Again, if mymailserver.com requires authentication, you need to set smtp_user and smtp_pass. > Thanks for your help! I check the logs but there isn't anything. When I sign up, nothing happens and I can't figure out why? Is Mailman running? If so, and there's nothing in logs, it is probably using some other (default) config and logging somewhere other than where you think. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

6 years, 6 months

[MM3-users] Re: Chinese characters in list name shows question mark (?)

by Mark Sapiro

On 10/23/19 7:12 AM, Franklin Weng wrote: > Hi list, > > I set up a mailman3 server and built a few lists to test. > > When I set the list name using Chinese characters, the web site shows > correctly. However in the subscribing welcome mail the list name was > shown in question marks. Screenshot is here: > > https://nextcloud.slat.org/index.php/s/CxPjpYMbQzQ6444 The name of the list from the screenshot seems to be zh-l10n(a)lists.linux.org.tw. The thing shown with ?s in the screenshot is the list's display_name (called Display name in Postorius). > The LANGUAGE_CODE settings.py (in Debian /etc/mailman3/mailman-web.py) > is set to zh-tw , and USE_I18N and USE_L10N are all set to True. These settings are all for Django and have nothing to do with how the list welcome message is sent. > Is there any other setting I should look at? The list's preferred_language setting (Postorius -> Settings -> List Identity -> Preferred Language. The list welcome message is encoded in Mailman's character set for the list's preferred_language. If this is English (USA), the character set is us-ascii. If it is set to Chinese (Taiwan), the character set is utf-8, and it should work. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

5 years, 9 months

[MM3-users] Re: About to Install MM3 -- Seeking clarification

by Mark Sapiro

On 10/22/22 21:29, Onyeibo Oku wrote: > > Now Mailman starts, and stops as desired when running as mailman user (both in > virtual and non-virtual environments). However, running the service via systemd > still fails. I now get a different traceback: > > Oct 23 03:43:00 mail.server.tld mailman[19410]: Traceback (most recent call last): ... > Oct 23 03:43:00 mail.server.tld mailman[19410]: with open(self._claimfile, 'w') as fp: > Oct 23 03:43:00 mail.server.tld mailman[19410]: PermissionError: [Errno 13] Permission denied: '/opt/mailman/mm/var/locks/mailman-cfg.lck|server.tld|19410|619270043958240354' > Oct 23 03:43:00 mail.server.tld systemd[1]: mailman3.service: Control process exited, code=exited, status=1/FAILURE What are ownership and permissions on /opt/mailman/mm/var/locks/ Assuming they are drwxr-xr-x mailman mailman it must be the case that systemd is not running the `/opt/mailman/venv/bin/mailman start` command as the `mailman` user. I suspect this is a Fedora issue. I note that https://docs.fedoraproject.org/en-US/quick-docs/understanding-and-administe… does not mention User or Group. You might try replacing ``` ExecStart=/opt/mailman/venv/bin/mailman start ExecReload=/opt/mailman/venv/bin/mailman restart ExecStop=/opt/mailman/venv/bin/mailman stop with ExecStart=sudo -u mailman /opt/mailman/venv/bin/mailman start ExecReload=sudo -u mailman /opt/mailman/venv/bin/mailman restart ExecStop=sudo -u mailman /opt/mailman/venv/bin/mailman stop ``` -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

2 years, 9 months

[MM3-users] Re: Flooded non-members with "prvs=<number>=<mail>"

by Mark Sapiro

On 12/22/24 04:35, Torge Riedel via Mailman-users wrote: > > I checked the non-members list of all my mailman3 lists and found a lot > entries like "prvs=<number>=<mail>", the <mail> is always an address > from @freenet.de > > A quick research mentions something related to bouncing and that such > mails should be blocked/filtered. But no further details. > > Can you give a hint where I need to change what to block such mails and > avoid that such entries floods the non-members list? If these are in fact bounces, the mta at freenet.de is sending bounce DSNs to the list address instead of (or possibly in addition to) the list-bounces address which is the envelope sender of the mail from Mailman. This behavior is wrong, but it is beyond your control. If you can identify a pattern in the From: address that would match these and only these mails, you could probably configure the incoming MTA to discard them. E.g., for Postfix, see https://www.postfix.org/header_checks.5.html You might think of using Mailman's Header filters to discard the messages, but this won't keep the nonmembers from being created as the creation happens before any of those checks. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

7 months, 1 week

[MM3-users] Re: Debian 12 postfix errors: "warning: connect to transport private/lmtp: Connection refused" and "status=deferred (mail transport unavailable)"

by Mark Sapiro

On 4/7/25 12:00, Philip Bondi wrote: > [mta] > incoming: mailman.mta.postfix.LMTP > outgoing: mailman.mta.deliver.deliver > lmtp_host: 127.0.0.1 > lmtp_port: 9024 Why are you setting this? Testing uses port 9024 to avoid conflicts, but the default port 8024 should be used for production. > [799] # postconf -n ... > virtual_mailbox_maps = hash:/opt/mailman/mm/var/data/postfix_lmtp Why do you have this? It may well be the issue. > ebill(a)systemdatabase.ca lmtp:[127.0.0.1]:9024 > ebill-bounces(a)systemdatabase.ca lmtp:[127.0.0.1]:9024 > ebill-confirm(a)systemdatabase.ca lmtp:[127.0.0.1]:9024 > ebill-join(a)systemdatabase.ca lmtp:[127.0.0.1]:9024 > ebill-leave(a)systemdatabase.ca lmtp:[127.0.0.1]:9024 > ebill-owner(a)systemdatabase.ca lmtp:[127.0.0.1]:9024 > ebill-request(a)systemdatabase.ca lmtp:[127.0.0.1]:9024 > ebill-subscribe(a)systemdatabase.ca lmtp:[127.0.0.1]:9024 > ebill-unsubscribe(a)systemdatabase.ca lmtp:[127.0.0.1]:9024 > > [797] # telnet localhost 9024 > Trying ::1... > Connection failed: Connection refused > Trying 127.0.0.1... > Connected to localhost. > Escape character is '^]'. > 220 shackleton12.systemdatabase.ca GNU Mailman LMTP runner 2.0 > QUIT > 221 Bye > Connection closed by foreign host. It looks like you can connect to port 9024 in this way. However I recommend removing `lmtp_port: 9024` from your mailman.cfg and rerunning `mailman aliases` to regenerate the postfix maps. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

3 months, 3 weeks

[MM3-users] Re: Invalid FilterAction: discard. Treating as discard

by Mark Sapiro

On 5/7/19 6:19 AM, Henrik Rasmussen wrote: > Having set the Filter content to Yes - regardless of other setting in the "Alter Messages" tab, make a HTML message come out empty. I can replicate this on my test list. I don't understand the question? Are you asking why the log contains "invalid FilterAction: discard. Treating as discard"? If so, that's a bug. See <https://gitlab.com/mailman/mailman/issues/592>. If you are asking why the message is empty after filtering and discarded, It is presumably because it is a text/html message as opposed to a multipart/alternative message with text/plain and text/html alternative parts. > I have tried debugging by changing each setting one-by-one and sending the user's mail, but no change makes the mail come out with a content. However setting "Filter content" to "no" makes the mail come out okay. There are list attributes filter_types and pass_types (also filter_extensions and pass_extensions, but those probably aren't relevant here) which affect this and which aren't exposed in Postorius. See <https://mailman.readthedocs.io/en/latest/src/mailman/handlers/docs/filterin…> for more info of content filtering. Presumably, the combination of filter_types and pass_types is not allowing text/html. You need to examine these settings in mailman shell, e.g. list(m.pass_types) -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

6 years, 2 months

[MM3-users] Re: Hyperkitty Recommended mySQL Database Setup & Archive Importing

by Mark Sapiro

On 12/8/20 9:05 AM, matthew(a)alberti.us wrote: > > From what I've been reading, it seems like Mailman3 expects > the mysql database to use UTF8 encoding. It appears, by default, that when > using the docker-mailman containers, that the character set is set to > latin1. This may explain why there are lots of import errors, and maybe also > why there are a lot of errors when threads are deleted in hyperkitty's UI. > > > > Is it the case that UTF8 is the preferred character set? I > also saw some threads indicating that maybe utf8mb4 is more preferred. I'd > like to try to change the character_set on the database, if that is > advisable. You want the MySQL encoding for the Mailman database to be utf8mb4. If not, as Ruth says, you will have issues with 4-bute utf-8 encodings. See <https://gitlab.com/mailman/mailman-suite/-/blob/master/mailman-suite_projec…> > Also, is Hyperkitty 1.3.4 officially released? I see that the > list.org instance is using it, but the latest announce I've seen is for > 1.3.3. No, it is not. lists.mailman3.org and mail.python.org both run the HEADs of the GitLab branches so they are generally ahead of the latest releases. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

4 years, 7 months

[MM3-users] Re: after upgrade: OperationalError: no such column: hyperkitty_mailinglist.archive_rendering_mode

by Mark Sapiro

On 1/31/24 08:02, Hans Mayer via Mailman-users wrote: > > > Dear All, > > recently I did a system os upgrade debian 11.8 -> 12.4 / bullseye -> > bookworm > > This did also upgrade Mailman3 version 3.3.3 -> 3.3.8 as a used the > packages which came with Debian OS. > > Initially after the upgrade I had some troubles with the MM3 services to > start, but I fixed this. > > Now almost everything is running. Only the Archives are not working. Issues with the Debian poackage should be reported to Debian. See <https://wiki.list.org/x/12812344>. > The browser shows a server error, when I look at the mailman-web.log I see > > ERROR 2024-01-31 15:14:14,403 2699563 django.request Internal Server > Error: /mailman3/hyperkitty/ > Traceback (most recent call last): > File "/usr/lib/python3/dist-packages/django/db/backends/utils.py", > line 84, in _execute > return self.cursor.execute(sql, params) > File > "/usr/lib/python3/dist-packages/django/db/backends/sqlite3/base.py", > line 423, in execute > return Database.Cursor.execute(self, query, params) > > sqlite3.OperationalError: no such column: > hyperkitty_mailinglist.archive_rendering_mode You need to run the Django admin (however that's spelled in the debian mailmanweb package) `migrate` command (and probably others, e.g., `collectstatic`, `compress`, `compilemessages`). -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

1 year, 6 months

[MM3-users] Re: Welcome message template errors

by Mark Sapiro

On 4/21/24 19:15, Mark wrote: > > And the site-owner receives an email with ... > > ==== > Internal Server Error: /mailman3/lists/ > DoesNotExist at /mailman3/lists/ > Site matching query does not exist. > Request Method: GET > Request URL: http://<REDACTED-PUBLIC-IPv4-ADDRESS>/mailman3/lists/ The above means you have to add <REDACTED-PUBLIC-IPv4-ADDRESS> to the list of ALLOWED_HOSTS in your Django settings, but see below. > Django Version: 4.2.11 > Python Executable: /opt/mailman/venv/bin/python3 > Python Version: 3.11.2 > ... etc ... > === > > Regarding the second line in the email - my thinking is that it should > be looking in "/opt/mailman/mm/var/lists" and not "/mailman3/lists/". It's talking about the path in the URL, not a path in the file system. > The error doesn't occur if the is no custom "Welcome message" template - > ie. the user is sent the default welcome message. I think you must be setting the custom template via Postorius and your Django setting for POSTORIUS_TEMPLATE_BASE_URL is wrong. It should be set to the scheme and host you use to access Postorius. I.e, if you normally access Postorius via something like https://example.com/mailman3 set POSTORIUS_TEMPLATE_BASE_URL = 'https://example.com' It is probably now set as POSTORIUS_TEMPLATE_BASE_URL = 'http://<REDACTED-PUBLIC-IPv4-ADDRESS>' -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

1 year, 3 months

[MM3-users] Re: templates

by Mark Sapiro

On 5/5/23 15:27, Christian via Mailman-users wrote: > Okay I sent a test message this morning and subscribers are reporting they > did not receive it. The smtp.log shows 'smtp to > degrowcali(a)lists.ccalternatives.org for 25 recips, completed in > 3276.275292634964 seconds' (see below) > I alone received it however and it is listed in the archive. Below is the > excerpt from smtp.log. Two of my 5 lists are exhibiting this behavior, while > 2 other lists are working normally. Please advise. Thank you. That message says Mailman delivered the message for 25 recipients to the outgoing MTA. If only one recipient received it, you first need to check the outgoing MTA's logs to see whether the message was delivered to the recipient's mail servers or not. Those logs should indicate that either the message was delivered to each recipient's first hop mail server or not. If not, there should be a reason why in the logs. If the message was delivered, it could have been rejected by an MTA further down stream in which case a DSN message should be returned, but possibly not for days, or it could have been delivered to the user's spam/bulk/junk/promo/whatever folder or silently discarded. Do your list's do bounce processing? If so, what's in Mailman's bounce.log? -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

2 years, 2 months

[MM3-users] Re: Problem with digests not being sent daily

by Mark Sapiro

On 6/21/24 22:04, Mark wrote: > > At ~04:50 I posted an email to the list. The message was delivered to > members and archived. > > Nothing showed up in /opt/mailman/mm/var/lists/<LISTID> Probably because (in Postorius) Settings->Digest->Enable Digests is No > Then, as "mailman", I ran "mailman digests --periodic" manually and got > the same error (as expected since there is nothing there) ... > > $ /opt/mailman/venv/bin/mailman digests --periodic > Traceback (most recent call last): ... > FileNotFoundError: [Errno 2] No such file or directory: > '/opt/mailman/mm/var/lists/<LISTID>/digest.mmdf' -> > '/opt/mailman/mm/var/lists/<LISTID>/digest.6.3.mmdf' This error should not occur if there is no digests.mmdf. It only does occur because you have set Digest size threshold to 0 under the assumption that 0 means unlimited which it doesn't. If you don't want digests sent on size, you need to set Digest size threshold to a very large number. I've created https://gitlab.com/mailman/mailman/-/issues/1158 to avoid throwing the exception, but setting the threshold to 0 is still wrong as it will cause a digest to be sent on every post. If the Enable Digests setting is Yes, then the lack of digests needs more investigation. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

1 year, 1 month

[MM3-users] Re: Unknown templates being used

by Mark Sapiro

On 2/14/21 5:44 AM, Tom Stoddard wrote: > I'm a list owner who has recently (within the last week) had my lists > migrated from mailman 2 to mailman 3. The UI that we are using is > Affinity. So you should be getting your support from EMWD. > I've reviewed all of the list templates and begun to customize > some of them. My issue is that I'm receiving messages that don't match > any of the templates. Here is an example of one that I just received > this morning. (The x's are inserted by me. The actual message provided > valid addresses.) > > The xxxxxxxxxx(a)xxxxxxxxxx.org list has 1 moderation requests waiting. > > > Held Subscriptions: > User:xxxxxxxxxx@gmail.com > > Please attend to this at your earliest convenience. That message is not built from a template. It comas from the `mailman notify` command which currently says ```> def _send_notice(mlist, count, detail): > """Creates and sends the notice to the list administrators.""" > subject = _('The {} list has {} moderation requests waiting.').format( > mlist.fqdn_listname, count) > # XXX This should be a template. > text = _("""The {} list has {} moderation requests waiting. ``` I.e., we know it should be a template, but it currently isn't. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

4 years, 5 months

[MM3-users] Re: Problems with New MM3 Installation

by Mark Sapiro

On 5/11/21 7:13 AM, mm3(a)steele.com wrote: > I have found work around for the settings issue but continue to have an issue with archiving new messages. New messages are stuck here; > > mailman3/mm/var/archives/prototype/[listname]@[domain]/new They aren't `stuck` there. That is where they are archived by the prototype archiver. This is separate from HyperKitty. > I have added this to my mailman.cfg: > > [archiver.hyperkitty] > class: mailman_hyperkitty.Archiver > enable: yes > configuration: /usr/local/mailman3/mailman-hyperkitty.cfg > > And created the "/usr/local/mailman3/mailman-hyperkitty.cfg” as such: > > base_url: http://localhost/archives/ > api_key: [key that matches “MAILMAN_ARCHIVER_KEY” in my settings.py] > > Some documentation says that the url should be "http://localhost/hyperkitty/ but url generates a page not found error on my setup while the archives are available at the “archives” url. > > But, now that I think about it, "localhost" doesn't make sense here because I am virtual hosts and "localhost" will probably just connect to the first virtual host rather than the mailmanweb virtual host. > > Any advice on what do do here? Set base_url: in /usr/local/mailman3/mailman-hyperkitty.cfg to a url that will reliably access HyperKitty, probably the same base URL that a user with a web browser would use. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

4 years, 2 months

[MM3-users] Re: can't get mail into mailman3 in docker

by Mark Sapiro

On 8/6/21 11:28 AM, bob B wrote: > I see nothing for today when the error happens > in /opt/mailman/var/logs/smtp.log > > But the log has a bunch of these, but these stopped around the 4th > > Aug 04 19:46:04 2021 (28) Available AUTH mechanisms: LOGIN(builtin) PLAIN(builtin) > Aug 04 19:46:04 2021 (28) Peer: ('127.0.0.1', 33912) > Aug 04 19:46:04 2021 (28) ('127.0.0.1', 33912) handling connection > Aug 04 19:46:04 2021 (28) ('127.0.0.1', 33912) EOF received > Aug 04 19:46:04 2021 (28) ('127.0.0.1', 33912) Connection lost during _handle_client() > Aug 04 19:46:04 2021 (28) ('127.0.0.1', 33912) connection lost These are normal messages from the startup of Mailman's lmtp runner. If you correlate these with mailman.log, you'll probably see the server starting at these times. The fact that there is nothing in this log correlating with the Postfix message 2021-08-06T09:08:41.974225-04:00 dockerhost postfix/lmtp[109783]: D630B3083ED0: to=<test(a)mailman-docker.###.####>, relay=localhost[127.0.0.1]:8024, delay=0.11, delays=0.05/0.02/0.04/0, dsn=4.4.2, status=deferred (lost connection with localhost[127.0.0.1] while receiving the initial server greeting) seems to say that Postfix isn't really connecting to Mailman's lmtp and this is consistent with netstat -l not showing something listening on port 8024. I note you have lmtp_host: localhost in the [mta] section in mailman-extra.cfg. Try changing that to lmtp_host: 127.0.0.1 -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

3 years, 11 months

[MM3-users] Re: Bulk update of members' moderation settings

by Mark Sapiro

On 11/18/21 12:37 PM, Alex King via Mailman-users wrote: > I imported a list from mailman 2.1 using > https://docs.mailman3.org/en/latest/migration.html. > > It happens that emergency moderation was set on the mailman 2.1 list, > and this has resulted in all members having Moderation set to Hold. I > now want to change that to Default Processing for all members. > > Is there a way to achieve that in bulk without clicking through the > members options for each member in Postorius? I don't understand how this happened. The mailman 2.1 list's emergency setting should hsave been imported as the emergency setting of the Mailman 3 list and should not affect individual member's moderation at all. However, to set the moderation action for all members of a list to default processing You can do something like this in mailman shell ``` bin/mailman shell -l list.example.com Welcome to the GNU Mailman shell Use commit() to commit changes. Use abort() to discard changes since the last commit. Exit with ctrl+D does an implicit commit() but exit() does not. The variable 'm' is the list.example.com mailing list >>> from mailman.interfaces.action import Action >>> for member in m.members.members: ... member.moderation_action = Action.defer ... >>> commit() ``` -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

3 years, 8 months

[MM3-users] Re: Alias Domains in postfix_domains wrong way around?

by Mark Sapiro

On 10/24/20 10:34 AM, Hans-Martin Mosner wrote: > > Is my understanding correct that postfix_domains should have the true (externally visible) domains as keys, and values > that basically don't matter, or should it have both the real domains and their alias names as keys? With such a > convoluted setup, my brain twists a little. I'm not sure what's going on in your case. I have two separate servers that use alias domains, and for both of them, the postfix_domains file lists the alias domain as the key and the real domain as the value. This works fine with the Postfix settings ``` transport_maps = hash:/path-to-mailman/var/data/postfix_lmtp relay_domains = hash:/path-to-mailman/var/data/postfix_domains virtual_alias_maps = hash:/path-to-mailman/var/data/postfix_vmap ``` Is this what you have? Note that the address should be remapped from list@true_domain to list@alias_domain by virtual_alias_maps prior to checking relay_domains. Perhaps you have some other unusual thing in your Postfix configuration. What does `postconf -n` report? This is what I see in the mail log for a typical delivery ``` Oct 18 13:30:44 example postfix/lmtp[13745]: 108CA1F8A03: to=<list(a)x.example.org>, orig_to=<list(a)example.org>, relay=127.0.0.1[127.0.0.1]:8024, delay=0.08, delays=0.05/0.01/0/0.02, dsn=2.0.0, status=sent (250 Ok) ``` where example.org is the true domain and x.example.org is the alias domain. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

4 years, 9 months

[MM3-users] Re: Installation -- error on mailman-web migrate

by Mark Sapiro

On 9/16/21 8:43 PM, Mark Dale via Mailman-users wrote: > > ... and then at the beginning of many errors printed to screen I now have... > > /opt/mailman/venv/lib/python3.9/site-packages/django_q/conf.py:139: UserWarning: Retry and timeout are misconfigured. Set retry larger than timeout, > failure to do so will cause the tasks to be retriggered before completion. > See https://django-q.readthedocs.io/en/latest/configure.html#retry for details. The current version of mailman_web/settings/mailman.py has ``` Q_CLUSTER = { 'retry': 360, 'timeout': 300, 'save_limit': 100, 'orm': 'default', } ``` Earlier versions omitted the `'retry': 360` which is the issue. > Completely not comprehending the above doc (and being absolutely clueless about django), I blindly added "retry: 30" ... > root@mx:~# nano -c /opt/mailman/venv/lib/python3.9/site-packages/django_q/tests/settings.py > > # Django Q specific > Q_CLUSTER = { > "name": "django_q_test", > "cpu_affinity": 1, > "testing": True, > "log_level": "DEBUG", > "django_redis": "default", > "retry": 30, Make that "retry": 360, > } > > > ... but no joy. Is it me or are there some pieces missing from the documentation? > > > _______________________________________________ > Mailman-users mailing list -- mailman-users(a)mailman3.org > To unsubscribe send an email to mailman-users-leave(a)mailman3.org > https://lists.mailman3.org/mailman3/lists/mailman-users.mailman3.org/ > -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

3 years, 10 months

[MM3-users] Re: Installing mailman 3

by Mark Sapiro

On 07/21/2017 09:36 AM, Rafael Mora wrote: > > Postfix <http://www.postfix.org/>: To setup Postfix, add the following > configuration to master.cf: > > # Support the default VERP delimiter.recipient_delimiter = > +unknown_local_recipient_reject_code = 550owner_request_special = > notransport_maps = > hash:/path-to-mailman/var/data/postfix_lmtplocal_recipient_maps = > hash:/path-to-mailman/var/data/postfix_lmtprelay_domains = > hash:/path-to-mailman/var/data/postfix_domains > > > I added that config at the bottom of the master.cf file and did a 'postfix > reload' and I got this error message: > > [root@localhost ~]# postfix reload > /usr/sbin/postconf: fatal: file /etc/postfix/master.cf: line 129: bad field > count > postfix/postfix-script: fatal: cannot execute /usr/sbin/postconf! First of all, in a report such as this, it would really help if you reported what is actually in master.cf in the vicinity of line 129. However, that said, the doc is wrong. The lines should be added to main.cf, not master.cf. See <https://gitlab.com/mailman/mailman-suite-doc/issues/14>. Also, what you post above is quite garbled. It is not clear what you meant, but the lines to add should be # Support the default VERP delimiter. recipient_delimiter = + unknown_local_recipient_reject_code = 550 owner_request_special = no transport_maps = hash:/path-to-mailman/var/data/postfix_lmtp local_recipient_maps = hash:/path-to-mailman/var/data/postfix_lmtp relay_domains = hash:/path-to-mailman/var/data/postfix_domains and should not be wrapped as in your post. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

8 years

[MM3-users] Re: Custom templates (was: Re: Re: nginx configuration on a multitasking server)

by Mark Sapiro

On 1/14/22 6:58 PM, David Newman wrote: > > On 1/14/22 6:35 PM, Mark Sapiro wrote: >> >> The place for feature requests for Mailman core (which this is) is >> https://gitlab.com/mailman/mailman/-/issues, however, this is not >> straightforward as Mailman core doesn't know if you even have >> Postorius or some other web management UI installed or what the URL to >> access it might be. Thus it's not just a simple matter of editing >> templates. It would also require a configuration setting for the base >> URL for the web admin UI and changes in logic to provide an >> appropriate replacement depending on whether or not the base URL was >> configured. > > Understood, thanks. Although this distinction is very important for > server admins, it's likely not meaningful for many list owners and > moderators. For them, MM2.1 sent an email saying "go here" and > MM3/Django/Postorius/Hyperkitty doesn't, at least not out of the box. > > There is some pending template in use by default, just not one with a > URL included. All the templates have defaults and the defaults don't have URLs because these are unknown to core. > Is this something better handled in the venv and/or Web installation > docs? Or should I file a feature request for core, or Postorious or > someplace else? I think a feature request for core is appropriate. I'm only saying implementing this is a non-trivial change. What I do on sites I admin is put custom versions of all these domain:admin:notice:new-list.txt list:admin:action:post.txt list:admin:action:subscribe.txt list:admin:action:unsubscribe.txt list:admin:notice:pending.txt list:member:digest:footer.txt list:member:digest:masthead.txt list:member:regular:footer.txt list:user:action:invite.txt list:user:action:subscribe.txt list:user:action:unsubscribe.txt list:user:notice:post.txt list:user:notice:probe.txt list:user:notice:welcome.txt in /opt/mailman/mm/var/templates/site/en and at mail.python.org which also has German language lists, in /opt/mailman/mm/var/templates/site/de. That could be documented, but it would be preferable to be able to document just a base web management URL and then put a replacement such as $web_ui in the default templates and replace that with a URL if configured or some generic text otherwise. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

3 years, 6 months

[MM3-users] Re: Header and footer templates in MM3

by Mark Sapiro

On 1/19/20 1:15 AM, Gila Halpern wrote: > I read the earlier message from Peter Mueller about message decoration, and > where the templates are in the file system. However, when I edit the > header and footer templates in Postorius, those templates do not appear in > the location specified. Should I be looking somewhere else? The templates created in Postorius are created in the postorius_emailtemplate table in Mailman's database and referenced via URLs like 'https://example.com/mailman3/api/templates/list/<list-id>/<template-name>'. I.e., they are not stored in the file system. > (I intend to > write a program to rotate though several headers/footers, so I need to know > where to have it write to.) If Postorius has set a template for a list, it has set the URI for that template and list to one similar to that above and that URI ultimately references the template in the database. You have choices. You can set the templates in Postorius and your program can do database queries to rotate the template text, or you can set the template URIs via Mailman's TemplateManager to file system paths or not set them and use the default paths (deleting them in Postorius will leave them unset). See <https://mailman.readthedocs.io/en/latest/src/mailman/rest/docs/templates.ht…> -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

5 years, 6 months

[MM3-users] Re: Mailman3 stopped working after Upgrade to Ubuntu 21.04

by Mark Sapiro

On 8/7/21 4:35 AM, Thomas Möhle via Mailman-users wrote: > Hey, > > I have set up mailman3 with mailman3-web on my mailserver and after > upgrading to Ubuntu 21.04 my setup stopped working. In what way? Do list posts get delivered? What works and what doesn't? What are the observable symptoms for things that don't work? > I can see repeating > log entries of mailman trying to reach something within hyperkitty and > getting an error. The same constant repeating cycle of log entries is > visible in my log files. It appears from your logs that Mailman core's attempts to reach the hyperkitty api may be failing. Mailman core uses the mailman_hyperkitty module to communicate with the hyperkitty api to get the url at which a message will be stored and to deliver a message to hyperkitty. This may not be working. There are probably unarchived messages in Mailman's var/archives/hyperkitty/spool/ directory. What do you get for example if you go in a web browser with hyperkitty running to https://mail.bitkomplex.de/mailman3/hyperkitty/api/mailman/urls?mlist=landw… And please don't post secret keys and change the value of api_key: in mailman-hyperkitty.cfg and the corresponding value in MAILMAN_ARCHIVER_KEY in your Django settings. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

3 years, 11 months

[MM3-users] Re: messages in queue/shunt

by Mark Sapiro

On 5/7/23 08:39, Christian via Mailman-users wrote: > Can/should old messages residing in queue/shunt be manually removed? I see > they were delivered. Is there any reason to leave them there? Thank you!! If they were delivered, they shouldn't be in queue/shunt/. Messages are shunted (placed in queue/shunt/) because of some exception in processing that caused processing to be aborted. It is possible for a message to be shunted even though it was delivered if the exception that caused it to be shunted occurred in the archive runner or the digest runner. When a message is shunted, an entry is made in mailman.log indicating the reason and giving a traceback. This information can be used to determine if there is some issue in Mailman that caused the exception, and if so, that issue can be fixed and the message's processing continued via `mailman unshunt`. Also, pointing `mailman qfile` at the shunted .pck file will not only display the raw message, but also metadata indicating which queue was being processed when the exception occurred. In any case, If the message is not going to be unshunted because it was spam or too old to be relevant or whatever, it should be removed from queue/shunt/ if for no other reason than so a future `mailman unshunt` won't requeue it. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

2 years, 2 months

[MM3-users] Re: Uncaught runner exception

by Mark Sapiro

On 4/8/20 1:17 PM, Wolfgang Bock via Mailman-users wrote: > > Checked the ownership .... > ls -la: > /var/lib/mailman3/cache > drwxrwx--- 3 root root 4096 Nov 9 22:27 61 > drwxrwx--- 3 root root 4096 Feb 3 17:37 6c > drwxrwx--- 3 root root 4096 Nov 12 11:59 71 > drwxrwx--- 3 root root 4096 Nov 6 12:01 72 > drwxrwx--- 3 root root 4096 Feb 5 21:52 7d > drwxrwx--- 3 root root 4096 Feb 5 23:05 8b > drwxrwx--- 3 root root 4096 Nov 12 23:41 94 > drwxrwx--- 3 root root 4096 Nov 7 08:31 ae > drwxrwx--- 3 root root 4096 Nov 12 23:40 fa It seems at some point Mailman core was run as 'root' rather than 'list'. ... > Changed the ownership for /var/lib/mailman3/cache to list:list and from that point my lists are responding. > > Same problem occurs before fort he list-directory, which I corrected some days ago. > > Is the ownship misconfiguration a coincidence or did I made a mistake somewhere?? This is a Debian/Ubuntu package question. I'm guessing but all of Mailman core and Mailman web should always be run as user:group list:list in your case. It appears that at some point these were started as root. Whether this is something you did wrong or an issue with whatever init/systemd scripts run these things, I can't say. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

5 years, 3 months

[MM3-users] Re: Mails are not archived at all -- how to check if Hyperkitty is running or not?

by Mark Sapiro

On 9/23/20 4:59 AM, cristian.bianchi(a)usi.ch wrote: > I have a similar problem with mailman 3.2.1 (debian package): > > # mailman info > GNU Mailman 3.2.1 (La Villa Strangiato) > Python 3.7.3 (default, Jul 25 2020, 13:03:44) > [GCC 8.3.0] > > the messages are not archived even if the key is set and hyperkitty is checked as active archiver. > > To analyze the problem I set DEBUG = True in /etc/mailman3/mailman-web.py. > In debug mode the problem does not occur, but this second problem appears: > https://gitlab.com/mailman/hyperkitty/-/issues/84 I don't understand. https://gitlab.com/mailman/hyperkitty/-/issues/84 is supposed to be fixed. Are you saying it isn't. > How can I fix it? Start by finding all the log messages relating to this archiving failure. If there are exceptions with tracebacks, what are they? Also, are there messages in var/archives/hyperkitty/spool/? If so, there will also be ''Exception in the HyperKitty archiver:' messages with an exception logged in mailman.log for each such message every time there is a list post. Note that HyperKitty is not a service that runs like Mailman core. The only check is can you access it via the web? -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

4 years, 10 months

[MM3-users] Re: How to add new lists?

by Mark Sapiro

On 11/2/22 15:21, fstltna--- via Mailman-users wrote: > I have been able to get the web interface up and the default "mailman@" address is there and appears in the list of lists, but when I add a list from the command line with a command like "mailman create foo(a)ourlists.gameplayer.club" it does not return a error and using "mailman lists" lists it as existing. However there is no new entries in the /etc/aliases file and the new list doesn't appear in the web interface... I am confused. How did you install Mailman. There is no default `mailman` address and Mailman doesn't create postfix mappings in /etc/aliases. If you have set ``` configuration: python:mailman.config.postfix ``` in the [mta] section of mailman.cfg, various mappings are created in Mailman's var/data directory? Configuration for Postfix is described at https://docs.mailman3.org/projects/mailman/en/latest/src/mailman/docs/mta.h…. There is also information there about other MTAs. As far as your list created with `mailman create` not appearing in the web UI, is the web UI communicating with the Mailman core instance in which you created the list. What does `mailman info` show and in particular is the host and port in its `REST root url` the same as that set in `MAILMAN_REST_API_URL` in Django's settings? -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

2 years, 8 months

[MM3-users] Re: Trouble Installing Mailman3

by Mark Sapiro

On 6/5/25 08:40, Henry Hartley via Mailman-users wrote: > > The other thing I had to do that was different to the instructions was to use pip to install setuptools. I got to a point where I had an error message that said "ERROR: Could not find a version that satisfies the requirement setuptools>=40.8.0" but when I ran "apt install python3-setuptools" it reported that it was "already the newest version (68.1.2-2ubuntu1.2)". However, running pip install setuptools (on a whim) installed version 80.9.0. After that "pip install mailman" worked. Perhaps I should have known this. I feel that making sure setuptools is properly installed might be something to add to the installation instructions. It depends how you create the venv. `virtualenv venv` will create the venv with `pip`, `setuptools` and `wheel` installed, but our install doc says do `python3 -m venv venv` and that only installs `pip` in the venv. I'll update the doc. > The next step in the instructions has me create /etc/mailman3/mailman.cfg. The first thing in that file is shown as: > > [paths.here] > var_dir: /opt/mailman/mm/var > > Looking in /opt/mailman, there is no mm directory. Should that have been created, or is that something I need to create myself? You need to create it. I'll update the doc for that too. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

1 month, 3 weeks

[MM3-users] Re: Templates

by Mark Sapiro

On 09/28/2017 01:56 PM, David Krantz wrote: > Hi again, > > I had the problem with getting two footers, i.e. two expansions of the > template. After some digging in the source code I found out that the > decorator made one expansion (without personalization) and that the > delivery handling made another (with personalization). So I hacked > pipelines/builtin.py with a new pipeline that excluded the decorate > step, recompiled and changed to that pipeline for my personalized > list. That worked. Nothing like a bit of shotgun debugging in > production. > > That solves my immediate problem. Is there a known and recommended way > of configuring personalized delivery so that you only get the > personalized footer? I have ideas, but I probably should share them > with the developer list instead. :-) This is clearly a bug and should be fixed. If you report this issue at <https://gitlab.com/mailman/mailman/issues>, we will try to fix it. > Of course this very nice functionality is a bit buried. I activated it > using the interactive python shell, seems less risky than changing the > database directly. There are many core settings not (yet) exposed in Postorius. This is something we need to fix, but it takes time and resources. In the mean time, 'mailman shell' is probably the best way to set them. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

7 years, 10 months

[MM3-users] Re: trying to clean up the "held messages" mess

by Mark Sapiro

On 5/28/21 12:37 AM, Bjoern Franke wrote: > Hi, > > I migrated a Mailman3 installation on Debian10 yesterday evening from > Debian packages to PIP (thanks to [1]). > > I think some bugs which existed before are now cleaned out, but mailman > still threw exceptions when trying to get the "held messages". What exceptions? Note that there is a lot of cleanup in the as yet unreleased core 3.3.5. You might consider installing that. To do so, in your virtualenv as the Mailman user do ``` pip install --upgrade git+https://gitlab.com/mailman/mailman ``` > The amount of held messages is now reduced, but e.g. a list with 120 > held messages before has till 59 held messages. And now I'm confused: > where are they? The held message entries are in the _request table with `request_type` = 1. For those entries, `key` is the message-id and `data_hash` is the `token` in the pended table and the `id` in the pended table is the `pended_id` of the related entries in the pendedkeyvalue table. The message itself is stored in the file system and is pointed to by the entry in the message table with the corresponding message-id. The exceptions you see are because entries are missing from the pended, pendedkeyvalue and/or message tables. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

4 years, 2 months

[MM3-users] Re: trying to clean up the "held messages" mess

by Mark Sapiro

On 5/28/21 11:32 AM, Bjoern Franke wrote: > > But since the upgrade from 3.2.1 to 3.3.x the list owners are spammed with > > ---- > some.listmember(a)domain.tld's subscription has been disabled on > nordwest(a)lists.ffnw.de due to their bounce score > exceeding the mailing list's bounce_score_threshold. This should really be better documented (sigh) Prior to Mailman 3.3.1, bounces were recorded but not processed. Beginning with Mailman 3.3.1, bounces are now properly processed. The issue is all those unprocessed bounces were still in the bounceevent table in the database and even though they were old and likely irrelevant, upon upgrade they were all processed resulting in many spurious disabling of delivery. This is disruptive, but is a one time thing and in your case is now done. For those who haven't yet upgraded, there are two ways to avoid this. The best way is before starting Mailman after upgrading, set ``` [mta] verp_probes: yes ``` in mailman.cfg to enable bounce probes. With probes enabled, users whose score reaches threshold are sent a probe message and their delivery is only disabled if the probe bounces which will only happen for currently undeliverable addresses. The other way is before starting Mailman after upgrading, run a database query similar to ``` UPDATE bounceevent SET processed = 1; ``` -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

4 years, 2 months

[MM3-users] Re: Disabled by bounce

by Mark Sapiro

On 1/31/23 22:32, Allan Hansen wrote: > Hi all, > > Apparently Google just started to get very aggressive about DKIM settings. Some hundreds of my subscribers have had their subscriptions disabled because of Google's bounces. > > I tried to 'Enable' some of them, expecting that their bounce scores would be reset, but when I list the users, the bounce scores are still 5. > > Please advise. Will enabling them without the bounce score being reset just cause the server to disable them again. Yes, Google has started enforcing a policy which is effectively DMARC p=reject even for senders whose domains don't publish DMARC p=reject or quarantine. To avoid this issue, you may need to set DMARC Mitigate unconditionally to Yes as well as DMARC mitigation action to Replace From: with list address if it isn't already. As far as enabling without resetting the bounce score is concerned, I think you are correct that they will just be disabled again on the next bounce. This is a bug and should be fixed. However, all of the above can be avoided without any DMARC changes by enabling bounce probes. in the [mta] section of mailman.cfg, set ``` verp_probes: yes ``` Since the probe is sent From: the list's domain, it should pass googles SPF and DKIM checks and not bounce. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

2 years, 6 months

Search results for query "sapiro"