Mailman 3 January 2021 - Mailman-Developers

before next release: disable backscatter in default installation
by Jo Rhett 08 Nov '22

08 Nov '22

Hi. There's a fairly simple problem here that needs to be addressed. And it's mostly a documentation/install problem. I'm hoping we can get this resolved before the next release. PROBLEM: Mailman comes out of the box ready to backscatter spam people. Yes, it's easy enough to fix. But because it comes stock this way, and is documented to install this way, most people install it to do this. Those of us who work in abuse departments are tired of hearing "well that's how Mailman works". We also object to having to teach people how to fix their mailman installations because it's not documented in the current manual. This is *exactly* like Sendmail 14 years ago. We didn't accept it then, and Sendmail fixed the problem. RESOLUTION: Mailman default installation should not backscatter in a default configuration. 1. Don't create backscatter aliases for subscribe/unsubscribe/etc by default. Nearly everyone uses web based signup. 2. Discard or hold messages from non-subscribers by default. I would think that it would be perfectly reasonable to have documentation on how to enable the 1980s-style -request / -subscribe etc aliases. However this documentation should have a note that this is against the AUP of nearly every network provider, and enabling it will likely cause them to get listed in various blacklists as a backscatter source. FYI: I know that this goes against the instincts of many old-time mailing list advocates here. But after dealing with a 10k/hour backscatter DoS my tolerance for this problem is understandably limited. Yes, it was a sweet day back in the 1980s. I was running a mailing list server and several UUCP gateways at the time, so I remember them well. But those days are past, and we need to deal with the reality of today. -- Jo Rhett Net Consonance : consonant endings by net philanthropy, open source and other randomness

21 168

Mailman, postfix and docker
by costavitorino＠gmail.com 28 Mar '21

28 Mar '21

Hi all I'm upgrading a mailman2 installation to mailman3. And I using a docker implementation, created by me and not Maxking's implementation. Everything is installed and is served correctly. I have created a postfix container that relays emails to a smtp server, and I can send emails by telnet from it. When I create a new thread in Hyperkitty, I receive an error from postfix container that the user (mailing list name) is not found. And I believe that is obvious because the aliases db are not accessible by postfix. So, my questions are: - Is there a way to share aliases db between containers(different hosts) of mailman and postfix? I thought about sharing volumes but I'm not quite sure if it's the best solution - I've seen: virtual_alias_maps = mysql:/etc/postfix/virtual/existing_aliases.cf, var/lib/mailman/data/virtual-mailman to add in postfix main.cf. Is it possible to do the same thing to aliases and transport_maps? And if so, how to set it in mailman configuration, because this is for mailman I've seen this thread: https://mail.python.org/archives/list/mailman-developers@python.org/thread/… And it seams kind or less the same situation. Does anyone have an idea or can give some pointers in how to accomplish this? Kind regards, Emanuel

5 7

GSoC 2021 ideas
by Terri Oda 15 Feb '21

15 Feb '21

Happy new year everyone! I'm starting to prep Python for GSoC 2021, and I wanted to issue the usual invitation that we'd be happy to have Mailman as a sub-org if anyone wants to do GSoC this year. (Mailman's been fine as a separate org, but I have to do the paperwork for Python anyhow so I figured I'd offer if there's interest but no one wants to cover the admin side of things.) Also, I have two mailman related ideas! 1. Seamless archive converter for 2.1 -> 3.* A friend was grousing that there was no "nice" way to retain old archive urls without keeping a static copy of things, and it got me to thinking that it would definitely be possible to build something that handled the old urls and redirected or made them work seamlessly, but we didn't do it because parsing the old archives basically requires you to scrape them because otherwise we couldn't guarantee that the urls would be stable from a re-parsed mbox. It's boring and finicky web scraping work to associate the old url and do the right things to make it work seamlessly in hyperkitty, but probably not too hard, so I was thinking that it might be suitable for a GSoC student. 2. Old mailman "skin" for postorius Make mailman look like the 2.1 interface for people who really love the old system. There's a few options that would be different, but the goal would be to make it pretty much look the same only with a few options changed, for people who are very change adverse. We had intended for it to be *possible* to reskin Postorius, but I don't think too many people have done it, so this would be a test to see how doable that is and probably fix any underlying issues that make reskinning the interface hard. Honestly, we could also have a student do a brand new skin if we had someone who loved UI design, but I suspect replicating the old interface would be less work, and since this year's GSoC hours have been cut in half, i'd rather start with something easier. I've been out of mailman dev for 3 years, so I'm probably not the ideal mentor, but I'm up for helping on either of those if Mailman as a whole is interested in doing the GSoC thing this year. Terri

5 10

Footer settings question: how common is dash-dash-space?
by Alessandro Vesely 02 Feb '21

02 Feb '21

Hi, my footer removal software fails on IETF's "last-call" mailing list because it looks for a line of underscores, whereas that list uses "-- ". That sequence is "the separator line between the body and the signature of a message", according to RFC 3676. However, I had never seen it in mailing list footers before. My experience with mailing lists is limited, so I ask you. Is that sequence common enough to try and catch it in a footer removal function? I asked the list owners, and they answered "We didn't choose these settings when the list was set up! It was probably just the default for that version of Mailman." Is that possible? TIA Ale --

6 18

which database is used in unittest?
by baguazhangba＠gmail.com 23 Jan '21

23 Jan '21

Hello All, Happy New Year 2021! I've configured mailman using sqlite as backend database. but when I run unittests, e.g. python3 -m nose2 -v mailman.model.test seems the data is not write to production sqlite database, so which database is used in unittest and how to check data in that db? thanks! Mike

3 7

Issues gathering list members
by costavitorino＠gmail.com 23 Jan '21

23 Jan '21

Hi I'm with mailman 3 in a custom docker environment. And I have all services running in independent containers, and they all communicate between them. I'm testing sending emails by telnet to a mailing list. But mailman don't send the mail to the mailing list members. But if I send the mailing list to xxxx-owners(a)xxxx.xx, all the owners receive the mailing list. Just when sending to the mailing list it doesn't work. If I access in shell the mailing list mailmanclient, and I get the members, it returns all members. I was debuging lmtp runner, and I can see that using "owners" it goes to queue "in", and Switchboard unqueue receives the message. But without "owners", it goes to "in" queue but unqueue method is not called. How can I debug why the mailing list is never sent to members? Another thing that is happening, is in Hyperkitty, the mailing list never shows as active. And it haves 0 participants and 0 discussions. When I'm creating a new thread, I receive the message that email was sent but it never arrives to any recipient. I thought that it requires an actual email to sent to turn the mailing list as active. But at this moment I'm not quite sure about it. Can you point me in the right direction to understand whats happening? kind regars, Emanuel

4 7

New release candidates for Postorius and Hyperkitty
by Abhilash Raj 17 Jan '21

17 Jan '21

Hey Everyone, Happy new year to everyone! I am pleased to announce new release candidates for Postorius and Hyperkitty along with new releases for Mailmanclient and Django-mailman3! These include both new features and bug fixes. Complete change log for each components: Postorius 1.3.4rc1: https://docs.mailman3.org/projects/postorius/en/latest/news.html Hyperkitty 1.3.4rc1 : https://docs.mailman3.org/projects/hyperkitty/en/latest/news.html Mailmanclient 3.3.2.: https://docs.mailman3.org/projects/mailmanclient/en/latest/src/mailmanclien… Django-mailman3 1.3.5: https://gitlab.com/mailman/django-mailman3/-/blob/master/README.rst These releases require Mailman Core 3.3.2 and Python 3.6+. Since these are pre-releases, you can install them using: $ pip install --pre --upgrade postorius hyperkitty Or you can also specify the exact versions: $ pip install --upgrade 'postorius==1.3.4rc1' 'hyperkitty==1.3.4rc1' Unless there is a release-blocker bug, stable versions for Postorius and Hyperkitty will be released in 2 weeks time. If you can test these releases, we encourage you to report any bugs to us on Gitlab. If you'd like to help out with translations for some of the new strings in your native language, the documentation is available here[1] on how to do that. [1]: https://docs.mailman3.org/en/latest/translation.html -- thanks, Abhilash Raj (maxking) On behalf of GNU Mailman Team

1 1

Logging
by costavitorino＠gmail.com 08 Jan '21

08 Jan '21

Hi Is it possible to log to console? I've changed gunicorn configuration: accesslog='-' errorlog='-' Then logs are not written to log files, but I can't see what's being logged. I'm on a docker environment and having log files it's something that I want to avoid to control disk usage. And I'm capturing what's shown in console, so it was great if I could log into console. Kind regards Emanuel

3 3

Is it possible to use Postorius alongside the legacy interface?
by Joe Corneli 01 Jan '21

01 Jan '21

Hello, I wonder if it's possible for users to choose between the new Postorius interface and the old text-based interface? Some users might prefer the text-based interface (e.g., for accessibility or just familiarity); others may prefer the new interface (which does look quite snazzy!). But, I didn't find any documentation about switching between the two. Joe

2 1