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

[MM3-users] Re: Trying to understand the mailman.cfg from the virtualenv installation guide

by Mark Sapiro

On 9/11/24 01:44, Danil Smirnov via Mailman-users wrote: > Hi Mark, > > ср, 11 сент. 2024 г., 06:26 Mark Sapiro <mark(a)msapiro.net>: > >> On 9/8/24 01:04, unite42--- via Mailman-users wrote: >>> [mta] >>> verp_confirmations: yes >>> verp_personalized_deliveries: yes >>> verp_delivery_interval: 1 >>> >>> What is the background of these recommendations? >> >> >> These settings enable VERP for all list messages. This results in more >> reliable bounce processing. >> > > > Is there any downsides of using it, so it's not enabled by default? With verp_delivery_interval set to 1, Mailman delivers a separate VERPed message to the MTA for each recipient whereas without this, Mailman delivers one or a few messages with multiple recipients each. This can have a performance impact. This is not a consideration for confirmations or personalized deliveries as those messages have only one recipient per message in any case. I.e., there is no performance impact from setting verp_confirmations: yes verp_personalized_deliveries: yes This is not the default because there can be issues with some MTAs dealing with the VERPed envelope sender. -- 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: Invalid FilterAction: discard. Treating as discard

by Mark Sapiro

On 5/8/19 4:45 AM, Henrik Rasmussen wrote: > Mark Sapiro wrote: > >> I'll post a follow-up with how to change pass_types via mailman shell >> when I work it out. > > Look forward to that, thanks. OK. in mailman shell with your list as 'm' >>> list(m.pass_types) ['multipart/mixed', 'multipart/alternative', 'text/plain'] >>> m.pass_types = ['multipart', 'text/plain', 'text/html'] >>> commit() >>> list(m.pass_types) ['multipart', 'text/plain', 'text/html'] >>> If you want to allow text from forwarded messages, add 'message/rfc822' in the pass_types list above. This setting will allow inspection of all elementary parts within any multipart parts ant will accept only text/plain and text/html. Then if collapse_alternatives is True (settable here or in Postorius), any multipart/alternative parts will by replaced by the first, presumably text/plain, sub-part. Finally, if convert_html_to_plaintext is True (settable here or in Postorius), any remaining text/html parts will be converted to plain text by lynx or whatever is configured as html_to_plain_text_command in mailman.cfg. This will ensure that filter_action is applied only to those messages which have no text/plain or text/html parts. -- 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: How to turn off Emergency Moderation for imported MM2 lists?

by Mark Sapiro

Mark Sapiro wrote: > On 7/28/20 5:40 AM, Brian Carpenter wrote: > > > > This is the client that I asking about in my post. > > She said what I > > did, did not work. This is what I did: > > > > Get into a venv as user mailman > > > > #mailman shell > > from mailman.model.mailinglist import MailingList > > mlist = MailingList('listname(a)listdomain.org') > > mlist.emergency = False > > commit() > > ctrl-d (just to be on the safe side) I half understand why this didn't work. Instantiating the list via ``` from mailman.model.mailinglist import MailingList mlist = MailingList('listname(a)listdomain.org') ``` for some reason (that's the half I don't understand) does not get the actual list object so changes you make to it don't affect the actual list. In `mailman shell` you need to do ``` mlist = getUtility(IListManager).get('listname(a)listdomain.org') ``` `getUtility` and `IListManager` are in the namespace in `mailman shell`, but outside of that, e.g. just in `python` in your virtualenv, you also need ``` from mailman.interfaces.listmanager import IListManager from zope.component import getUtility ``` -- 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: Fwd: [Django] ERROR (EXTERNAL IP): Internal Server Error: /archives/search

by Mark Sapiro

On 1/23/23 12:17, Odhiambo Washington wrote: > On Mon, Jan 23, 2023, 23:09 Mark Sapiro <mark(a)msapiro.net> wrote: >> >> It's actually trying to create fulltext_index, so clearly it's not >> looking at the fulltext_index that you are. If you are following >> https://docs.mailman3.org/en/latest/install/virtualenv.html you probably >> need to specify the path in /etc/mailman3/settings.py as in, e.g. >> ``` >> HAYSTACK_CONNECTIONS['default']['PATH'] = >> '/opt/mailman/mm/var/fulltext_index' >> > > Doesn't this follow the same rules of getting the settings.py from gitlab? I don't understand. Can you be more specific? Note that when following https://docs.mailman3.org/en/latest/install/virtualenv.html you create /etc/mailman3/settings.py and it contains ``` from mailman_web.settings.base import * from mailman_web.settings.mailman import * ``` to establish a base set of settings that can be augmented or overridden in /etc/mailman3/settings.py. mailman_web.settings.mailman in turn contains ``` HAYSTACK_CONNECTIONS = { 'default': { 'ENGINE': 'haystack.backends.whoosh_backend.WhooshEngine', 'PATH': "fulltext_index", ``` which is a relative path for fulltext_index and may need to be set to an absolute path. -- 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: Uncaught runner exception: 'str' object has no attribute 'decode'

by Mark Sapiro

On 09/29/2017 02:10 AM, Henrik Rasmussen wrote: > Thanks, Mark. > > @Abhilash, any chance for an update to Docker-mailman to fix these known issues soon? > https://gitlab.com/mailman/mailman/issues/256 @Henrik, Contrary to what I said earlier, your issue is not exactly #256. #256 is fixed in Mailman to the extent that the defective message no longer breaks Postorius or causes exceptions in REST, but the underlying issue causing the "KeyError: 'content-transfer-encoding'" exception is not in Mailman. It's in the Python email package. The issue that needs to be fixed for that is <https://bugs.python.org/issue27321>. The bottom line is at this point, <https://bugs.python.org/issue27321> no longer affects Postorius, but it still affects message delivery. I have opened <https://gitlab.com/mailman/mailman/issues/406> about defending against that and also the UnicodeEncodeError you included in your OP. > @Mark, > > Mark Sapiro wrote: >>> >>> bin/mailman queue >>> var/queue/shunt/1506287488.0529234+cbf8a31cad265a40a18e8955be0c9ad8b48a1b66.pck > > Isn't this a Mailman 2 command? What is the Mailman 3 equivalent? No. 'bin/mailman queue' is the Mailman 3 command. The Mailman 2 command is 'bin/show_qfiles'. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

9 years

[MM3-users] Re: mailman import21

by Mark Sapiro

On 10/11/20 8:26 AM, Brian Carpenter wrote: > > On 10/11/20 11:06 AM, Mark Sapiro wrote: >> Also, things like message and digest headers and footers that are MM 2.1 >> list attributes that are templates in MM 3, although import21 tries to >> handle those. > > So are how these templates showing up on my Mailman 3 servers? They show > up in /opt/mailman/mm/var/templates/lists/listname/en/. I have a client > whose old footer template from Mailman 2 still showing up in list > messages. He then created a custom footer template via Affinity (same > thing happen in Postorius) and both custom footers were being included > in list messages. As I said, (msg|digest)_(header|footer) are MM 2.1 list attributes, not templates. Those, and I think only those, are converted to var/templates/lists/listname/lc/ templates by import21. If I understand you correctly, after creating a custom template in Postorius or Affinity, both that and the old template were included in delivered posts. I don't understand that. For Postorius at least, when a custom template is created, it sets the URI for the template in core to point to Postorius to get the template. the old template in var/templates/lists/listname/lc/ should not be referenced. -- 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: lists.mailman3.org web site.

by Mark Sapiro

On 08/27/2016 09:15 PM, Mark Sapiro wrote: > I have just upgraded the lists.mailman3.org Mailman 3 installation to > the heads of the GitLab mailman* branches. A major change is the switch > from Persona to Django Allauth. This requires some configuration which > isn't yet complete. Hopefully it will all be working very soon. > > In the mean time if you want to log in to Hyperkitty/Postorius you > should be able to log in with your email address. The first time you do > this, Mailman probably won't even know your password so you'll have to > request a reset. Once you have reset your password, you should be able > to log in. > > I hope to have several social auth choices working very soon. Update: Django Allauth is now properly configured and working on lists.mailman3.org. You can log in by email/password as noted above or via OAuth2 with a GitHub, GitLab, Google or Facebook identity. If you are an existing user, in order to use one of the OAuth2 providers, you must first log in with email and password and go to the Account Connections tab in your account settings and add the 3rd party account(s) you wish. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

10 years, 1 month

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

by Mark Sapiro

On 4/15/19 10:16 AM, Dmitry Makovey wrote: > So our mailman got stuck again, but I did follow your advise. Here's > what I've found: > > On 3/29/19 2:58 PM, Mark Sapiro wrote: >> >> If the out runner is running, is there a file named *.bak along with all >> the *.pck files in queue/out/ > > there is one, yes: > > # ls -l /srv/mail/mailman/core/var/queue/out/*.bak > -rw-rw---- 1 100 65533 21512 Apr 14 19:45 > /srv/mail/mailman/core/var/queue/out/1555296331.8386078+2c905825b82708e5a682b872b0aff320c9dd0b2e.bak > > what does it indicate and what should my steps be to resolve this (other > than restarting queues)? This says that the out runner has picked up that entry and is processing it, but, of course, it is now hung somehow in that process. Examining that entry with 'mailman qfile' may possibly provide a clue, but I suspect not as when you restart the out runner process it will reprocess that entry (which is the whole purpose of saving it as .bak during processing), and this time it seems to go OK. You might try running strace or similar on the procees to try to see what it is waiting on. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

7 years, 6 months

[MM3-users] Re: Can't subscribe on new site/domain

by Mark Sapiro

On 10/11/25 4:58 AM, Flo via Mailman-users wrote: > On 10/11/25 04:45, Mark Sapiro wrote: >> >> The fields in order are: >> >> 1. the id of the django_mailman3_maildomain entry >> 2. the mail_domain in the django_mailman3_maildomain entry >> 3. the site_id the django_mailman3_maildomain entry >> 4. the id of the django-site entry >> 5. the domain name in the django-site entry >> 6. the display name in the django-site entry >> > > For me it looks like that 5 is the display name and 6 the domain name Possibly sqlite displays them in a different order. To be sure you could do SELECT mail_domain, domain FROM django_mailman3_maildomain JOIN django_site ON site_id = django_site.id; which will display only the mail_domain and the site domain. > For each of the lines mail_domain (field 2) is the same as the domain > name (field 6, if I am right). > Any domain name (and mail_domain, which has the same value) differs from > the ones in any other line. > > So they are unique. Anyway, if all 3 mail_domain values are unique, that doesn't explain the issue. I can only guess that this is a Debian package issue. -- 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: Display the last post time for a list or all lists.

by Mark Sapiro

Mark Sapiro wrote: > It wouldn't be difficult to do, but at the moment, it wouldn't provide anything useful because of <https://gitlab.com/mailman/mailman/issues/453>. The issue is now fixed in the branch head at least. The script corresponding to https://www.msapiro.net/scripts/last_post.py would be almost identical except for changing real_name to display_name, last_post_time to last_post_at and fixing the print for Python 3. Also changing the string interpolation to format(), it becomes: --------------------------------------------------- import time def last_post(mlist): print('{}: Last post at {}'.format( mlist.display_name, time.ctime(mlist.last_post_at) )) --------------------------------------------------- The only issue with the script is for a list which has had no posts since #453 was fixed, mlist.last_post_at is None and time.ctime(None) is the current date and time. That too could easily be fixed like: --------------------------------------------------- import time def last_post(mlist): if mlist.last_post_at: print('{}: Last post at {}'.format( mlist.display_name, time.ctime(mlist.last_post_at) )) else: print('{}: No posts'.format(mlist.display_name)) --------------------------------------------------- The trickier part is the Mailman 3 mailman withlist command has no option to run for all lists, so you'd need to run it something like #!/bin/sh for list in `mailman lists -q`; do mailman withlist -r last_post -l $list done -- 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

Search results for query "sapiro"