Mailman-coders November 2011

mailman-coders@python.org

17 participants
90 discussions

[Bug 885715] [NEW] Mailman should add a Received header
by Barry Warsaw 23 May '15

23 May '15

Public bug reported: When the LMTP runner or any other injection vector receives a message, it should add a Received header to indicate the time at which Mailman got the message. Use case: there is a spam filter between the incoming MTA and Mailman, and the message sits in the spam filter for a while before its administrator accepts the message. ** Affects: mailman Importance: Undecided Status: New ** Tags: mailman3 ** Tags added: mailman3 -- You received this bug notification because you are a member of Mailman Coders, which is subscribed to GNU Mailman. https://bugs.launchpad.net/bugs/885715 Title: Mailman should add a Received header To manage notifications about this bug go to: https://bugs.launchpad.net/mailman/+bug/885715/+subscriptions

2 3

[Bug 869347] [NEW] bin/test script fails unless zope.testrunner is pip installed
by Stephen A. Goss 23 May '15

23 May '15

Public bug reported: I am running Mailman 3 in a virtualenv with most of its package dependencies installed through pip. In this particular configuration, if I run `buildout`, the test script that is created cannot seem to load zope.testrunner properly: $ bin/test -vv Traceback (most recent call last): File "bin/test", line 35, in <module> import zope.testrunner ImportError: No module named testrunner The workaround is to remove the eggs folder, pip install zope.testrunner, and then re-run buildout. ** Affects: mailman Importance: Undecided Status: New ** Tags: mailman3 -- You received this bug notification because you are a member of Mailman Coders, which is subscribed to GNU Mailman. https://bugs.launchpad.net/bugs/869347 Title: bin/test script fails unless zope.testrunner is pip installed To manage notifications about this bug go to: https://bugs.launchpad.net/mailman/+bug/869347/+subscriptions

3 3

[Bug 837700] [NEW] Issues with users subscribed via their preferred address
by Barry Warsaw 23 May '15

23 May '15

Public bug reported: You can subscribe an address to a mailing list, and you can subscribe a user to a mailing list, if the user has a preferred address. There are two problems with this. * You can actually end up getting doubly subscribed (i.e. same address with the same role). Normally, the subscription mechanism prevents you from subscribing the same address with the same role more than once, or the same user with the same role more than once. However, if you subscribe once by the address, and once by the user whose preferred address is that same address, the double subscription occurs. * The Member repr() doesn't distinguish between a member that's subscribed via explicit address, and one that's subscribed via preferred address. This means that the above will appear as a double address subscription, when it's really not. Are these substantial problems that need to be fixed? Almost certainly the second one does, but I haven't come up with an elegant repr() for the case of a user subscribed with their preferred address. ** Affects: mailman Importance: Low Status: Confirmed ** Tags: mailman3 -- You received this bug notification because you are a member of Mailman Coders, which is subscribed to GNU Mailman. https://bugs.launchpad.net/bugs/837700 Title: Issues with users subscribed via their preferred address To manage notifications about this bug go to: https://bugs.launchpad.net/mailman/+bug/837700/+subscriptions

2 2

[Bug 793669] [NEW] feature request. change the 'list my other subscriptions' to also report if the other subscriptions are enabled or disabled
by Laura Creighton 23 May '15

23 May '15

Public bug reported: I have a large number of subscriptions to mailing lists on python.org. Delivery is disabled on most of them. This is exactly how I want it -- I don't want to keep up with the day to day matters of the list, but I do want to be able to re-anable delivery and discuss things there should the need arise without having to create a new account. However, when I go on vacation, what I want to do is also disable a few high volume lists that I normally want to read every day. Thus what I would really like to see is which lists are currently enabled, so I can decide which one(s) if any I want to disable for the vacation. I don't want to disable the low-volume ones. Given that I am going to have to re-enable the lists I want to read when I get back one by one, I want the number of lists that get temporarily disabled to be as small as possible. And I want to be able to email me the short list of what they were so I can remember which ones to turn on again. Interestingly, the text of the message indicates a usability misunderstanding, and a grammatical error. The text reads: You can view a list of all the other mailing lists at python.org for which you are a member. Use this if you want to make the same membership option changes to this other subscriptions. It should be 'these other subscriptions'. And the people who want to make the same membership option changes to all the lists they are subscribed to on a site are precisely the users who don't need the feature. They can blindly go off and change things globally. If the number of lists they are subscribed to on that site is 1 -- i.e. they do not have any other subscriptions -- then saying to do things globally will do no harm. The reason you might want to look is to check because you might _not_ want to do things globally. And once you are dealing with that case, things you do not want to do globally, then of course you want to see the current settings that you have. But for my use case, all I care about is changing delivery status. All the other options I either never change, or want to change globally. I suspect that most people use mailing lists in this way, where the only thing they need to worry about is changes due to vacation, but I haven't conducted any polls to test this theory. This request sent to the tracker because the automated mail I recevied when some moderator refused to accept my mail to mailman-dev with this feature request, with the message 'why don't you just disable delivery globally and then reset them globally when you return' said that this was the place for feature requests and not the mailing list. Hope this request is clearer and explains why globally disabling is not what I want to do. Thank you ** Affects: mailman Importance: Undecided Status: New -- You received this bug notification because you are a member of Mailman Coders, which is subscribed to GNU Mailman. https://bugs.launchpad.net/bugs/793669 Title: feature request. change the 'list my other subscriptions' to also report if the other subscriptions are enabled or disabled

4 9

[Bug 707295] [NEW] Membership Management Better Search Function
by arky 23 May '15

23 May '15

Public bug reported: In Mailman Membership Management, I would like to option to filter search results by status viz (hide,nomail [reason], ack, not metoo, nodupes, digest, plain) ** Affects: mailman Importance: Undecided Status: New -- You received this bug notification because you are a member of Mailman Coders, which is subscribed to GNU Mailman. https://bugs.launchpad.net/bugs/707295 Title: Membership Management Better Search Function

4 3

[Bug 514625] [NEW] Mailman 3 member functions should "guess" address membership
by Stephen Turnbull 23 May '15

23 May '15

Public bug reported: Feature request Mailman 3 should try to fuzzily match unknown author addresses against the member database. For example, if "turnbull(a)sk.domain.tld" is subscribed, then "turnbull(a)shako.sk.domain.tld" probably is the same person. This could probably be efficiently implemented by storing domains in bigendian order, and testing for prefixes. Bonus points for matching on full name. For example, for me personally firstname(a)xemacs.org and lastname(a)xemacs.org both resolve to me, which could not be reasonably associated by algorithm, but I always use the same full name, which would be strong indication that it's me. Possibly matching on mailbox would also be useful, but only as an optimization to be confirmed by one of the other two (or maybe if the match is unique?) Note that for the first example, the second "turnbull@shako" variant is today used only by spammers; I haven't sent a message with that "From" since 1995, and I doubt it's been used as "Sender" more than a very small handful of times since 2000. For this reason, the feature should *never* be automated. Instead, it should be part of the moderation interface. If a match is detected, then an option should appear saying something like "This address appears to be an alias for <member name and address>. [ ] Add <unknown address> as an alias for <member>." ** Affects: mailman Importance: Undecided Status: New ** Description changed: Feature request Mailman 3 should try to fuzzily match unknown author addresses against the member database. For example, if "turnbull(a)sk.tsukuba.ac.jp" is subscribed, then "turnbull(a)shako.sk.tsukuba.ac.jp" probably is the same person. This could probably be efficiently implemented by storing domains in bigendian order, and testing for prefixes. Bonus points for matching on full name. For example, for me personally firstname(a)xemacs.org and lastname(a)xemacs.org both resolve to me, which could not be reasonably associated by algorithm, but I always use the same full name, which would be strong indication that it's me. Possibly matching on mailbox would also be useful, but only as an optimization to be confirmed by one of the other two (or maybe if the match is unique?) Note that for the first example, the second "turnbull@shako" variant is today used only by spammers; I haven't sent a message with that "From" since 1995, and I doubt it's been used as "Sender" more than a very small handful of times since 2000. For this reason, the feature should *never* be automated. Instead, it should be part of the moderation interface. If a match is detected, then an option should appear saying something like "This address appears to be an alias for <member name and - address>. [ ] Add <unknown> address as an alias for <member>." + address>. [ ] Add <unknown address> as an alias for <member>." ** Description changed: Feature request Mailman 3 should try to fuzzily match unknown author addresses against - the member database. For example, if "turnbull(a)sk.tsukuba.ac.jp" is - subscribed, then "turnbull(a)shako.sk.tsukuba.ac.jp" probably is the same + the member database. For example, if "turnbull(a)sk.domain.tld" is + subscribed, then "turnbull(a)shako.sk.domain.tld" probably is the same person. This could probably be efficiently implemented by storing domains in bigendian order, and testing for prefixes. Bonus points for matching on full name. For example, for me personally firstname(a)xemacs.org and lastname(a)xemacs.org both resolve to me, which could not be reasonably associated by algorithm, but I always use the same full name, which would be strong indication that it's me. Possibly matching on mailbox would also be useful, but only as an optimization to be confirmed by one of the other two (or maybe if the match is unique?) Note that for the first example, the second "turnbull@shako" variant is today used only by spammers; I haven't sent a message with that "From" since 1995, and I doubt it's been used as "Sender" more than a very small handful of times since 2000. For this reason, the feature should *never* be automated. Instead, it should be part of the moderation interface. If a match is detected, then an option should appear saying something like "This address appears to be an alias for <member name and address>. [ ] Add <unknown address> as an alias for <member>." -- Mailman 3 member functions should "guess" address membership https://bugs.launchpad.net/bugs/514625 You received this bug notification because you are a member of Mailman Coders, which is subscribed to GNU Mailman.

4 5

[Bug 881320] [NEW] Mailman 3 translations
by Vincent Fretin 23 May '15

23 May '15

Public bug reported: In mailman 3.0.0a8, the locales directory contains actually mailman 2 translations. The unsubscribe mail is not translated into French because the msgids are different between mailman2 and mailman3. Do you have a script to generate a pot file and sync the po files? I had to generate myself the French mo file from the po. It's maybe a good idea to include mo files in the sdist, or generate mo files at mailman startup. ** Affects: mailman Importance: Undecided Status: New -- You received this bug notification because you are a member of Mailman Coders, which is subscribed to GNU Mailman. https://bugs.launchpad.net/bugs/881320 Title: Mailman 3 translations To manage notifications about this bug go to: https://bugs.launchpad.net/mailman/+bug/881320/+subscriptions

4 5

[Bug 881316] [NEW] RuntimeError: maximum recursion depth exceeded while pickling an object
by Vincent Fretin 23 May '15

23 May '15

Public bug reported: With mailman 3.0.0a8. I got the following traceback from a html mail with html signature+image from Outlook, Oct 14 09:21:31 2011 (17579) Traceback (most recent call last): File "/home/zope/mailman/src/mailman/core/runner.py", line 138, in _one_iteration self._process_one_file(msg, msgdata) File "/home/zope/mailman/src/mailman/core/runner.py", line 220, in _process_one_file keepqueued = self._dispose(mlist, msg, msgdata) File "/home/zope/mailman/src/mailman/runners/incoming.py", line 64, in _dispose process(mlist, msg, msgdata, mlist.start_chain) File "/home/zope/mailman/src/mailman/core/chains.py", line 90, in process link.function(mlist, msg, msgdata) File "/home/zope/mailman/src/mailman/chains/hold.py", line 245, in _process nmsg.send(mlist, **dict(tomoderators=True)) File "/home/zope/mailman/src/mailman/email/message.py", line 198, in send self._enqueue(mlist, **_kws) File "/home/zope/mailman/src/mailman/email/message.py", line 216, in _enqueue virginq.enqueue(self, **str_keywords) File "/home/zope/mailman/src/mailman/core/switchboard.py", line 133, in enqueue msgsave = cPickle.dumps(_msg, protocol) RuntimeError: maximum recursion depth exceeded while pickling an object It's really an obscure issue for me. I resolved it by adding sys.setrecursionlimit(10000) in src/mailman/core/switchboard.py before msgsave = cPickle.dumps(_msg, protocol) line 134 and doing bin/mailman unshunt fixed the issue. By default sys.getrecursionlimit() returns 1000. If you want to have a test email to reproduce, I can ask my customer to send an email to a test mailing-list. ** Affects: mailman Importance: Undecided Status: New -- You received this bug notification because you are a member of Mailman Coders, which is subscribed to GNU Mailman. https://bugs.launchpad.net/bugs/881316 Title: RuntimeError: maximum recursion depth exceeded while pickling an object To manage notifications about this bug go to: https://bugs.launchpad.net/mailman/+bug/881316/+subscriptions

4 6

[Bug 881312] [NEW] Properly decode payload
by Vincent Fretin 23 May '15

23 May '15

Public bug reported: With mailman 3.0.0a8 in src/mailman/runners/command.py, there is an issue with body = part.get_payload(decode=True) if you send a mail to test-unsubscribe(a)example.com, with accents in the mail, in the signature for example, you get a body string encoded in iso-8859-1. When you try to concatenate iso-8859-1 string with unicode, this the case when the mail body with "Unprocessed:" and "Ignored:" is generated, you get a UnicodeDecodeError ascii. The code should be modified to correctly decode the mail body like it is done in src/mailman/pipeline/decorate.py mcset = msg.get_content_charset() or 'us-ascii' unicode(msg.get_payload(decode=True), mcset) There are similar issues in src/mailman/rules/administrivia.py with the line: if line == '': where it give you a warning because it compare an iso to an empty unicode. and in src/mailman/rules/approved.py where it simply crashes: Oct 18 17:54:41 2011 (18207) Traceback (most recent call last): File "/home/zope/mailman/src/mailman/core/runner.py", line 138, in _one_iteration self._process_one_file(msg, msgdata) File "/home/zope/mailman/src/mailman/core/runner.py", line 220, in _process_one_file keepqueued = self._dispose(mlist, msg, msgdata) File "/home/zope/mailman/src/mailman/runners/incoming.py", line 65, in _dispose process(mlist, msg, msgdata, mlist.start_chain) File "/home/zope/mailman/src/mailman/core/chains.py", line 68, in process if link.rule.check(mlist, msg, msgdata): File "/home/zope/mailman/src/mailman/rules/approved.py", line 81, in check if ':' in line: UnicodeDecodeError: 'ascii' codec can't decode byte 0xe0 in position 14: invalid continuation byte Currently, mailman is unusable for another language than English, like French. ** Affects: mailman Importance: Undecided Status: New -- You received this bug notification because you are a member of Mailman Coders, which is subscribed to GNU Mailman. https://bugs.launchpad.net/bugs/881312 Title: Properly decode payload To manage notifications about this bug go to: https://bugs.launchpad.net/mailman/+bug/881312/+subscriptions

4 6

[Bug 826861] [NEW] doctests failing on Centos 6 due to "unprintable" error objects
by Stephen A. Goss 23 May '15

23 May '15

Public bug reported: On Centos 6 (64 bit), the following tests fail due to what looks like some kind of Python object printing quirk with the version that ships with RHEL 6 and derivatives: src/mailman/model/docs/membership.txt src/mailman/model/docs/users.txt The errors all look like this: File "/home/sgoss/mailman_clone/src/mailman/model/docs/membership.txt", line 296, in membership.txt Failed example: gwen_member.address = new_address Differences (ndiff with -expected +actual): Traceback (most recent call last): - ... - UnverifiedAddressError: gperson(a)example.com + File "/usr/lib64/python2.6/doctest.py", line 1248, in __run + compileflags, 1) in test.globs + File "<doctest membership.txt[68]>", line 1, in <module> + gwen_member.address = new_address + File "/home/sgoss/mailman_clone/src/mailman/model/member.py", line 116, in address + raise UnverifiedAddressError(new_address) + UnverifiedAddressError: <unprintable UnverifiedAddressError object> The system reports Python version 2.6.5. This error does not seem to occur with Python 2.7 on Ubuntu 11.04. One possibility is that for some reason when it comes to printing out those error objects, Python doesn't have a default encoding scheme for representing the embedded unicode strings in that particular context (email address in the above example). ** Affects: mailman Importance: Undecided Status: New ** Tags: mailman3 -- You received this bug notification because you are a member of Mailman Coders, which is subscribed to GNU Mailman. https://bugs.launchpad.net/bugs/826861 Title: doctests failing on Centos 6 due to "unprintable" error objects To manage notifications about this bug go to: https://bugs.launchpad.net/mailman/+bug/826861/+subscriptions

3 7

← Newer
1
2
3
4
5
6
7
8
9
Older →

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

Mailman-coders November 2011