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

[MM3-users] Re: Reset bounce count in MM3

by iMark

On 2026-03-09 05:03, Mark Sapiro wrote: > It's much easier to use mailman shell as in > ``` > $ 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 >>>> mbr = m.members.get_member('user(a)example.com') >>>> mbr.bounce_score = 0 >>>> commit() > ``` > But what is the point of doing this? If the member is bouncing, the > score will continue to increment, and if not the bounces will become > stale and a future bounce will set the score to 1. ================================================ From time to time Microsoft will spit the dummy and start blocking mail from the lists. This will last a day or two, but in the meantime bounce counts get cranked up which can send some lists over the edge and start disabling/removing subscribers. In a bash script I can query the the database like so: UPDATE MEMBER SET bounce_score = 3 FROM address WHERE address.id = member.address_id AND email LIKE'%hotmail.com' AND member.bounce_score >= 4 AND member.list_id = '<LISTID>' The idea is that the script can be croned hourly for a couple of days on multiple lists to wind down the bounce_counts (preventing disable/remove) while the Microsoft issue gets resolved. > You mentioned reset_bounce.py, but that doesn't reset scores. It sets > delivery enabled for users whose delivery is disabled by bounce, ... Ahh, I see. Thank you for setting me straight on that.

2 months, 2 weeks

[MM3-users] Re: using SSH/TLS with external MTA

by Roland Giesler

On 2024/07/27 22:58, Odhiambo Washington via Mailman-users wrote: > On Sat, Jul 27, 2024 at 11:34 PM Mark Sapiro <mark(a)msapiro.net> wrote: > >> On 7/27/24 13:05, Odhiambo Washington via Mailman-users wrote: >>> ASMTP works on the premise of "we can relay mail for you if we can >>> authenticate you with certain credentials" >>> Adding the IP to "hosts we can relay mail for" negates the need to >>> authenticate, IMHO. >>> So I am still waiting for how MM3 does ASMTP. Does it? >> >> If by ASMTP, you mean SMTP AUTH, Mailman does it if smtp_user: and >> smtp_pass: are provided in mailman.cfg. >> >> >> I agree that an authenticated user should be able to submit mail to be >> relayed. Just guessing here, but perhaps the issue is the OP is >> submitting to the deprecated port 465 instead of 587. >> > Port 465 is not quite depreciated. It's the decision of the Admin to decide > to listen on/and accept SMTP AUTH on which port. > On the servers I run, and based on Exim as the MTA, I accept SMTP AUTH On > ports 465, 587 and 2525 (this is meant to subvert some ISPs!). > But let's await Roland's response. The server runs power-mailinabox, which uses Postfix indeed. All my mail, and quite a few others' mail is sent from the using SMTP-Auth on port 465. Actually, this message is sent via that server too. So it's clear to me that MM3 is not authenticating for some reason. > >

1 year, 10 months

[MM3-users] Re: Help needed to create filters that were working in Mailman 2.

by Mark London

Mark - I'm an idiot. I previously had modified some of the text in Postorius to be shorter, and somehow had deleted the Accept feature! I hope not to ever have to upgrade our email server ever again. It took much years off my life. In any event, here's one of the text that I modified to be briefer. Perhaps you can see why I shortened it. Besides which, who goes on vacation these days, and disables emails from any source? I have users that literally have tens of thousands of unread email. They don't bother to delete it. Grrr... - Mark help_text=_( 'Set this option to Enabled to receive messages posted to this ' 'mailing list. Set it to Disabled if you want to stay subscribed, ' "but don't want mail delivered to you for a while (e.g. you're " "going on vacation). If you disable mail delivery, don't forget " 'to re-enable it when you come back; it will not be automatically ' 're-enabled.' On 6/21/2023 4:21 PM, Mark Sapiro wrote: > On 6/21/23 11:36 AM, mrl(a)psfc.mit.edu wrote: >> I'm using version 1.3.8 Which I installed on Ubuntu 22 on April >> 18. I'm guessing this feature was installed in 1.3.9? > > The attached screenshot (which may not make it to the list but should > be in your direct copy) is from Postorius 1.3.8 installed via pip. If > you don't see this, possibly you have installed the Debian/Ubuntu > package and this is a Debian issue. >

2 years, 11 months

[MM3-users] Re: Hiding an ML from a site

by Odhiambo Washington

On Thu, Feb 8, 2024 at 1:21 AM Mark Sapiro <mark(a)msapiro.net> wrote: > On February 7, 2024 1:34:32 PM PST, Odhiambo Washington < > odhiambo(a)gmail.com> wrote: > > >What should I do then? Edit some file? Apply some patch? > >It's pretty easy for me to create a DNS record for the same server (IP) > for > >use with the web host that I'd like to hide. > > > In the Domains view in postorius, you need separate mail hosts. Then you > need to change the domain of one of the lists to the 'new' mail host. This > implies changing its address which may not be acceptable. > Looking at Postorius, they do have different mail hosts. See the screenshot at: https://imgur.com/a/GsiiEjP Or, you can just live with it as is. > The problem with this is that those who go to the site have been trying to subscribe to the other ML. Note, I have been without power at my house for almost 3 days now, so I > don't have access to all my usual resources. > In Kenya, the blackouts can last equally long at times - and there is nothing we can do about it because the power distributor is a monopoly. Sorry for all the loss you've incurred from the foods getting spoilt in the fridge :-( -- Best regards, Odhiambo WASHINGTON, Nairobi,KE +254 7 3200 0004/+254 7 2274 3223 In an Internet failure case, the #1 suspect is a constant: DNS. "Oh, the cruft.", egrep -v '^$|^.*#' ¯\_(ツ)_/¯ :-) [How to ask smart questions: http://www.catb.org/~esr/faqs/smart-questions.html]

2 years, 3 months

[MM3-users] [Errno 13] Permission denied: '/var/lock/mailman/mailman-cfg

by David Newman

Greetings. I am attempting a complete Mailman 3 installation on a Debian 10 system that already runs Mailman 2, following this documentation: https://docs.mailman3.org/en/latest/install/virtualenv.html Even following this guide, I'd had a lot of issues with permissions, where directories and files were owned by list rather than mailman. After correcting those issues, I then got to the point of checking mailman3's status, and systemctl shows this error: Nov 19 20:09:57 [REDACTED] mailman[1463]: Error in atexit._run_exitfuncs: Nov 19 20:09:57 [REDACTED] mailman[1463]: Traceback (most recent call last): Nov 19 20:09:57 [REDACTED] mailman[1463]: File "/usr/lib/python3.7/logging/__init__.py", line 2038, in shutdown Nov 19 20:09:57 [REDACTED] mailman[1463]: h.flush() Nov 19 20:09:57 [REDACTED] mailman[1463]: File "/opt/mailman/venv/lib/python3.7/site-packages/mailman/core/logging.py", line 67, in flush Nov 19 20:09:57 [REDACTED] mailman[1463]: if self._stream: Nov 19 20:09:57 [REDACTED] mailman[1463]: AttributeError: 'ReopenableFileHandler' object has no attribute '_stream' There is an October 2019 thread on this error [1] but I think it's in a different context, where the OP already had mailman3 running, and was trying to migrate Mailman 2 lists. Also, regarding those permissions issues above, it matters which "layout" value is in /etc/mailman3/mailman.cfg. I'd tried "dev" because I saw a post from Mark Sapiro recommending that, but then mailman3 wanted me to create a bunch of directories under /etc, even though $var_dir was set to /var/lib/mailman3. Things seem to go better, except for the confusion between lists/mailman ownership, when I set layout to "debian", which was the default. Thanks in advance for clues on addressing the ReopenableFileHandler error above. dn [1] https://lists.mailman3.org/archives/list/mailman-users@mailman3.org/thread/…

4 years, 6 months

[MM3-users] Re: Things I miss about mailman2

by Quanah Gibson-Mount

--On Wednesday, March 18, 2020 2:58 AM -0700 Mark Sapiro <mark(a)msapiro.net> wrote: > This was implemented by > <https://gitlab.com/mailman/postorius/-/merge_requests/397> and released > in Postorius 1.3.0. Ok, that would be the issue... Debian 10 only has 1.2.4-1 bundled. I'll see if I can find a backport for 1.3. >>> This was not "removed". It just hasn't been implemented in MM 3 yet. >> >> So in terms of available functionality that was present in MM2, it was >> removed. It may need to be re-implemented, but this was a core feature >> that was present in the prior version that is no longer available. > > > Yes, we are aware of many things that are not yet exposed/implemented in > Postorius and elsewhere. There are currently 580 open issues at > <https://gitlab.com/groups/mailman/-/issues>. We are working on these > things as we can. We would love to have your help in fixing these. In > the mean time, there are really only 2 or 3 core developers working on > this and we are all volunteers, so it takes time. I fully understand and would love to contribute. The problem being that I already am doing this for multiple other projects where we're facing the same issue -- Lots of people use the software and there are the occasional contributions, but few people who are willing to be core members to really help drive the projects, and I can only drive so many and do my paying job. :/ Regards, Quanah -- Quanah Gibson-Mount Product Architect Symas Corporation Packaged, certified, and supported LDAP solutions powered by OpenLDAP: <http://www.symas.com>

6 years, 2 months

[MM3-users] Re: No e-mail distribution after import of archive

by Odhiambo Washington

On Tue, Nov 12, 2024 at 10:34 AM Markus Grandpré < markus.grandpre(a)uni-konstanz.de> wrote: > Dear Mr. Sapiro, > > on our test system we noticed the problem that after importing a mailing > list archive from a Mailman system v2.1 into a Mailman system v3.3.8, > e-mails are no longer distributed via this mailing list. E-mails that > are written to the list find their way into the archive, but they are > not sent to any member of the list. > > Before the archive was imported, writing an e-mail to the list and > distributing this e-mail via the list was not a problem. Unfortunately, > we cannot find any indication of this issue in the log of the mailman > system v3.3.8. Please help us to further analyze and solve this problem, > as we are planning to import 188 archives into our new productive > mailman system v3.3.8 during the next months. > Archives are imported into HyperKitty and should not in any way interfere with Mailman Core. From what I learnt recently, you can even import an archive into HyperKitty without having a ML in Core associated with it. So my starting point would be to look at mailman.log and see what goes in there. I could even up the debugging level if need be. For now, is there anything abnormal in mailman.log or smtp.log? -- Best regards, Odhiambo WASHINGTON, Nairobi,KE +254 7 3200 0004/+254 7 2274 3223 In an Internet failure case, the #1 suspect is a constant: DNS. "Oh, the cruft.", egrep -v '^$|^.*#' ¯\_(ツ)_/¯ :-) [How to ask smart questions: http://www.catb.org/~esr/faqs/smart-questions.html]

1 year, 6 months

[MM3-users] Re: hyperkitty: "incorrect string value" for botched utf-8 encoding

by thoralf

hi Mark, thank you for your answer. On 5/27/20 8:30 PM, Mark Sapiro wrote: […] >> (1366, "Incorrect string value: '\\xF0\\x9F\\x99\\x82\\x0AA...' […] >> Content-Type: text/plain; charset="utf-8" >> Content-Transfer-Encoding: quoted-printable […] >> kann halt NIE am mittwoch. viel spass =F0=9F=99=82 […] >> Content-Type: text/html; charset="utf-8" >> Content-Transfer-Encoding: quoted-printable […] >> <div dir=3D=22auto=22>kann halt NIE am mittwoch. viel spass >> =F0=9F=99=82<= > When HyperKitty gets an exception like this in archiving a message, it > queues it in /var/lib/mailman3/archives/hyperkitty/spool and on every > new post, retries the queued messages. This message fails again on retry > and is requeued. > > How would you suggest this be handled more gracefully? the way mailman3 is handling exceptions is totally fine and couldn't possibly be more graceful - i was referring to the error itself: > The underlying issue has been seen before. The basic problem is in MySQL > and MariaDB the encoding is utf8 which only accepts utf8 encodings up to > 3 bytes. It needs to be utf8mb4 to accept characters outside the base plane. okay, i thought that the html part in the offending mail should've been 🙂 and that this was the culprit - thanks for clarifying this. I've checked - database, table and columns were utf8mb4 already, whereas the connection to the db apparently not. Adding 'charset': 'utf8mb4' to the options-dict in the databases-variable in /etc/mailman3/mailman-web.py fixed the issue - django passes this on to mysqlclient. case closed :) thank you very much & with kind regards, thoralf.

5 years, 11 months

[MM3-users] Re: Moderated unsub requests are not provided by mailman3 service

by Joachim Schrod

Mark Sapiro writes: > On 11/29/24 4:47 PM, Joachim Schrod wrote: > > > System information: mailman3 3.3.3; Postorius (mailman3-web) 1.3.4; > > both on Debian 11 (bullseye). > > > > I know that this is out of date; thus a definitive info "this is fixed > > by now" is quite OK for me. (Debian upgrade is planned as the next > > action after getting Mailman3 to work for that use case.) > > Yes, this is fixed. Thanks for the information that it's fixed in Postorius 1.3.6; that information is a great help, it means that I don't need to spend more time on debugging resolved issues. I tried to gather data if an update to Debian 12 (bookworm) would help. But packages.debian.org is not helpful at all, for me. I know that this is beyond the realm of this mailing list, but perhaps someone here has that information and can supply it: Debian 12 (bookworm) package mailman3-web has the version "0+20200530-2.1". This is supposed to be from some source package mailman-suite that is a salsa project that I don't grok at 1st and 2nd sight. Can anybody please provide me with information what versions of mailman3 core and mailman3-web are in this package; i.e., in Debian 12 (bookworm)? Sorry, but I'm not versed enough to understand the information about Debian packages that my search delivers. Cheers, Joachim -- =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Joachim Schrod Email: joachim(a)schrod.org Roedermark, Germany ``How do we persuade new users that spreading fonts across the page like peanut butter across hot toast is not necessarily the route to typographic excellence?'' -- Peter Flynn

1 year, 5 months

[MM3-users] Re: Handling of (un)subscribe mails for already (un)subscribed users

by Andreas Barth

* Mark Sapiro (mark(a)msapiro.net) [201210 22:51]: > On 12/10/20 11:51 AM, Andreas Barth wrote: > > Hi, > > > > I noticed that a mail to -subscribe for an already subscribed user > > doesn't get any response (about line 225 in > > mailman/src/mailman/app/subscriptions.py). Same but different for > > unsubscribe for users not on the list. > > > I tested and I see this, so there is a bug, but I don't know what the > issue is. Ok, if you consider this a bug it makes sense to fix it upstream and not just locally. > The code you point to should raise AlreadySubscribedError and > <https://gitlab.com/mailman/mailman/-/blob/master/src/mailman/commands/eml_m…> > should catch it and report, but this doesn't seem to be happening.. Actually I don't see where it should send the mail in the code you refered to, but I'm pretty new to mm3-code and the involved coding guidelines. I would tend to add something along msg = UserNotification( self.mlist.owner_address, self.mlist.owner_address, subject, text, self.mlist.preferred_language) msg.send(self.mlist) (with appropriate subject, ...) before https://gitlab.com/mailman/mailman/-/blob/master/src/mailman/app/subscripti… but of course that could be the wrong approach (but as subscriptions.py is sending out mail at other places, that's my first idea). > These things should be fixed, but it's more involved than just a patch > to fix the error. For all but trivial fixes, the workflow should be > something along the lines of the following. [...] Thanks, I'll try to have a look into that. Regards, Andi

5 years, 5 months

Search results for query "sapiro"