Mailman 3 August 2013 - Mailman-coders

[Bug 1060951] [NEW] Bug getting i18n'ed attachment filenames (RFC2231)
by Aurélien Bompard Dec. 30, 2014

Dec. 30, 2014

Public bug reported: RFC 2231 allows filenames to have non-ascii characters. The get_filename() method in Python's Message class handles this by calling email.utils.collapse_rfc2231_value() at the end of get_filename. This method returns the filename in Unicode. This fails in Mailman because the mailman.email.message.Message class has a wrapper around get() and __getitem__() to return unicode headers. As a result, the collapse_rfc2231_value() tries to transforms into unicode an already unicode string, and I get the following exception: File "/usr/lib/python2.7/email/utils.py", line 319, in collapse_rfc2231_value return unicode(rawval, charset, errors) TypeError: decoding Unicode is not supported A possible solution to this would be to make Mailman's Message get_filename() method be more than just an exception-catching wrapper, and re-implement the original get_filename() method, inserting a conversion to str before calling collapse_rfc2231_value(). Does this make sense ? Any other idea for a possible solution ? ** 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/1060951 Title: Bug getting i18n'ed attachment filenames (RFC2231) To manage notifications about this bug go to: https://bugs.launchpad.net/mailman/+bug/1060951/+subscriptions

3 8

[Bug 1157854] [NEW] test_not_enough_parameters hangs on OS X
by Barry Warsaw Nov. 30, 2014

Nov. 30, 2014

Public bug reported: test_not_enough_parameters (mailman.rest.tests.test_users.TestLogin) This test hangs completely on OS X 10.6.8 and 10.8.2. It does not hang on Ubuntu 13.04 AFAICT. Commenting out this test allows the rest of the test suite to run to completion, although I get three test failures (on OS X 10.8.2). test_cannot_create_duplicate_list (mailman.rest.tests.test_lists.TestLists) test_cannot_create_list_in_missing_domain (mailman.rest.tests.test_lists.TestLists) test_existing_user_error (mailman.rest.tests.test_users.TestUsers) Further, C-c leaves the RESTRunner running. ** 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/1157854 Title: test_not_enough_parameters hangs on OS X To manage notifications about this bug go to: https://bugs.launchpad.net/mailman/+bug/1157854/+subscriptions

1 3

[Bug 880613] [NEW] PostgreSQL constraints disabled
by Barry Warsaw Nov. 29, 2014

Nov. 29, 2014

Public bug reported: In src/mailman/database/sql/postgres.sql, several foreign key constraints have been commented out because otherwise the test suite, or other aspects do not work. This is probably because there are cases where Mailman violates these constraints, but SQLite3 (the default database) does not care. All disabled constraints are marked with XXX in the above file. Really, these should be fixed. ** Affects: mailman Importance: Medium 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/880613 Title: PostgreSQL constraints disabled To manage notifications about this bug go to: https://bugs.launchpad.net/mailman/+bug/880613/+subscriptions

1 2

[Bug 1091586] [NEW] Frequent OperationalError: database is locked with sqlite
by Aurélien Bompard Nov. 26, 2014

Nov. 26, 2014

Public bug reported: I'm seeing frequent locking errors with SQLite on the current version of mailman3 from bzr. Here's an example traceback : Traceback (most recent call last): File "/srv/dev/mailman/bin/runner", line 9, in <module> load_entry_point('mailman==3.0.0b2-', 'console_scripts', 'runner')() File "/srv/dev/mailman/mailman3/mailman/src/mailman/bin/runner.py", line 253, in main runner.run() File "/srv/dev/mailman/mailman3/mailman/src/mailman/core/runner.py", line 96, in run filecnt = self._one_iteration() File "/srv/dev/mailman/mailman3/mailman/src/mailman/core/runner.py", line 173, in _one_iteration config.db.commit() File "/srv/dev/mailman/mailman3/mailman/src/mailman/database/base.py", line 77, in commit self.store.commit() File "/srv/dev/mailman/lib/python2.6/site-packages/storm-0.19-py2.6-linux-x86_64.egg/storm/store.py", line 122, in commit self._connection.commit() File "/srv/dev/mailman/lib/python2.6/site-packages/storm-0.19-py2.6-linux-x86_64.egg/storm/databases/sqlite.py", line 126, in commit self.raw_execute("COMMIT", _end=True) File "/srv/dev/mailman/lib/python2.6/site-packages/storm-0.19-py2.6-linux-x86_64.egg/storm/databases/sqlite.py", line 154, in raw_execute return Connection.raw_execute(self, statement, params) File "/srv/dev/mailman/lib/python2.6/site-packages/storm-0.19-py2.6-linux-x86_64.egg/storm/database.py", line 322, in raw_execute self._check_disconnect(raw_cursor.execute, *args) File "/srv/dev/mailman/lib/python2.6/site-packages/storm-0.19-py2.6-linux-x86_64.egg/storm/database.py", line 371, in _check_disconnect return function(*args, **kwargs) sqlite3.OperationalError: database is locked It happens during the normal processing of Mailman, not while using the API (to my knowledge), so it seems different from bug #856004. I can reproduce this easily by letting Mailman work for some time. After the locking error, new messages are stuck in var/queue/pipeline: $ ls var/queue/* var/queue/archive: var/queue/bad: var/queue/bounces: var/queue/command: var/queue/digest: var/queue/in: var/queue/lmtp: var/queue/nntp: var/queue/out: var/queue/pipeline: 1355781960.9870529+18f0acdf07aab616f793ecf76acfa1994e62b3a9.pck 1355799347.787621+213f3a6a971717ad1ba27f3d56b65767747a45da.pck 1355814973.8623891+4e66eeadef9b8c9ae99e4ed022c6c93f5af75c05.pck 1355782349.546113+ee14cb323574b647d9014f742e3f505e388d553e.pck 1355800543.460628+c455625c1ad86a5a77c9c515fcb6d555ba439b19.pck 1355815808.8860879+ca3fb5f45999e47c92756489daa951672ba4304d.pck 1355792739.8787439+19973fbcf763f036655aa0a7d477231246afdfbd.pck 1355802301.7830889+7989a3199c7eb491b65176937c0c1658cca9d46d.pck 1355820018.1813829+5005ded04da3574a7f3cd7472c9d635c577b6c32.pck 1355794764.597765+6cf379ad16c948117bb65f0aed0b3c230268ffd2.pck 1355803681.7455051+52d185cdcaecaf883bd43f375f5253567f5cd568.pck 1355822188.0491021+6a3cab9831d03b4f94475f745fa508c5126116b4.pck 1355794889.8323901+e14e25e342ec59709266040d830d7e12be2a1261.pck 1355805362.052228+422e573a6117edb7aae1acf9150ac58c210cedba.pck 1355822293.2363911+627f13dd2121bcb0f1916bc0448cc3b8adc9ad1f.pck 1355795937.2442579+62193d05ee80c6a22213cf2ff211c135fb7d38d9.pck 1355807374.9844501+8de57ab53819fac59d64705bcc56f86899d22634.pck 1355798662.8335559+554da4e7690e116e6fabee5764ac69829dc25685.pck 1355808273.143533+ed1acbf580c596ed921e062ece575877f0c9304b.pck var/queue/rest: var/queue/retry: var/queue/shunt: 1355776302.1302171+c657867b33bf6e37554ff83ee26f4096a77c1114.pck var/queue/virgin: And I need to restart mailman to make it process them again. The locked caused the message to be shunted. I'll attach the mailman.log file, where you'll see many "database is locked" errors, happening on different requests. ** 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/1091586 Title: Frequent OperationalError: database is locked with sqlite To manage notifications about this bug go to: https://bugs.launchpad.net/mailman/+bug/1091586/+subscriptions

2 8

[Bug 1082746] [NEW] Automated processes can swamp a list with web subscription requests.
by Mark Sapiro Oct. 2, 2014

Oct. 2, 2014

Public bug reported: There are discussions of this in threads at <http://mail.python.org/pipermail/mailman- users/2012-October/074213.html>, <http://mail.python.org/pipermail /mailman-users/2012-October/074278.html> and <http://mail.python.org/pipermail/mailman- users/2012-November/074412.html>. The Mailman developers do not think there is any way to prevent this other that disabling web subscribe entirely, as by definition, subscription requests come from unauthenticated users. However, an attempt will be made to mitigate this by making a site option to include a dynamically generated hidden hash in the subscribe form which will at least require an automated process to first GET and parse the listinfo form immediately prior to POSTing it. ** Affects: mailman Importance: Medium Assignee: Mark Sapiro (msapiro) Status: In Progress -- You received this bug notification because you are a member of Mailman Coders, which is subscribed to GNU Mailman. https://bugs.launchpad.net/bugs/1082746 Title: Automated processes can swamp a list with web subscription requests. To manage notifications about this bug go to: https://bugs.launchpad.net/mailman/+bug/1082746/+subscriptions

3 8

[Bug 1095422] [NEW] Non-queue runners should not create var/queue/* directories
by Barry Warsaw April 22, 2014

April 22, 2014

Public bug reported: The LMTP and REST runners are not queue runners. They are daemon processes which do not scan directories. They should not create directories in var/queue, probably by disabling the creation of the Switchboard in their subclass. ** Affects: mailman Importance: Low Status: Triaged ** 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/1095422 Title: Non-queue runners should not create var/queue/* directories To manage notifications about this bug go to: https://bugs.launchpad.net/mailman/+bug/1095422/+subscriptions

2 8

[Bug 518517] [NEW] Add the equivalent of postconf(1)
by Barry Warsaw April 22, 2014

April 22, 2014

Public bug reported: Postfix has a very handy command called postconf(1) which lets you dump all or one Postfix configuration variable. It would be very handy for Mailman 3 to have something similar. ** Affects: mailman Importance: Medium Assignee: Barry Warsaw (barry) Status: Triaged ** Affects: mailman/3.0 Importance: Medium Assignee: Barry Warsaw (barry) Status: Triaged ** Also affects: mailman/3.0 Importance: Medium Assignee: Barry Warsaw (barry) Status: Triaged -- Add the equivalent of postconf(1) https://bugs.launchpad.net/bugs/518517 You received this bug notification because you are a member of Mailman Coders, which is subscribed to GNU Mailman.

5 19

[Bug 1203519] [NEW] hide_address cannot be set through REST
by Barry Warsaw April 22, 2014

April 22, 2014

Public bug reported: This is missing: === modified file 'src/mailman/rest/preferences.py' --- src/mailman/rest/preferences.py 2013-01-01 14:05:42 +0000 +++ src/mailman/rest/preferences.py 2013-07-21 15:10:10 +0000 @@ -86,6 +86,7 @@ acknowledge_posts=GetterSetter(as_boolean), delivery_mode=GetterSetter(enum_validator(DeliveryMode)), delivery_status=GetterSetter(enum_validator(DeliveryStatus)), + hide_address=GetterSetter(as_boolean), preferred_language=GetterSetter(language_validator), receive_list_copy=GetterSetter(as_boolean), receive_own_postings=GetterSetter(as_boolean), ** Affects: mailman Importance: High Status: Triaged ** 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/1203519 Title: hide_address cannot be set through REST To manage notifications about this bug go to: https://bugs.launchpad.net/mailman/+bug/1203519/+subscriptions

1 2

[Bug 1184376] [NEW] REST server crashes on "reopen"
by Aurélien Bompard April 22, 2014

April 22, 2014

Public bug reported: Mailman's REST server crashes on the "mailman reopen" command. Here are the only lines I get in the logs: May 26 12:29:01 2013 (30830) Master started May 26 12:29:05 2013 (30844) RESTRunner runner started. May 26 12:29:05 2013 (30844) Starting REST server May 26 12:29:05 2013 (30837) DigestRunner runner started. May 26 12:29:05 2013 (30843) RetryRunner runner started. May 26 12:29:05 2013 (30848) CommandRunner runner started. May 26 12:29:05 2013 (30845) BounceRunner runner started. May 26 12:29:05 2013 (30846) LMTPRunner runner started. May 26 12:29:05 2013 (30850) NNTPRunner runner started. May 26 12:29:05 2013 (30838) PipelineRunner runner started. May 26 12:29:06 2013 (30839) ArchiveRunner runner started. May 26 12:29:06 2013 (30849) VirginRunner runner started. May 26 12:29:06 2013 (30836) IncomingRunner runner started. May 26 12:29:06 2013 (30840) OutgoingRunner runner started. May 26 12:29:33 2013 (30830) Master watcher caught SIGHUP. Re-opening log files. May 26 12:29:33 2013 (30848) CommandRunner runner caught SIGHUP. Reopening logs. May 26 12:29:33 2013 (30843) RetryRunner runner caught SIGHUP. Reopening logs. May 26 12:29:33 2013 (30838) PipelineRunner runner caught SIGHUP. Reopening logs. May 26 12:29:33 2013 (30837) DigestRunner runner caught SIGHUP. Reopening logs. May 26 12:29:33 2013 (30839) ArchiveRunner runner caught SIGHUP. Reopening logs. May 26 12:29:33 2013 (30836) IncomingRunner runner caught SIGHUP. Reopening logs. May 26 12:29:33 2013 (30846) LMTPRunner runner caught SIGHUP. Reopening logs. May 26 12:29:33 2013 (30849) VirginRunner runner caught SIGHUP. Reopening logs. May 26 12:29:33 2013 (30845) BounceRunner runner caught SIGHUP. Reopening logs. May 26 12:29:33 2013 (30850) NNTPRunner runner caught SIGHUP. Reopening logs. May 26 12:29:33 2013 (30840) OutgoingRunner runner caught SIGHUP. Reopening logs. As you can see in the "caught SIGHUP" lines, there is nothing about the REST server, and indeed there is no "runner --runner=rest:0:1" process running anymore. I have to restart mailman to get the REST server back online. ** Affects: mailman Importance: Undecided Status: New ** Tags: mailman3 rest -- You received this bug notification because you are a member of Mailman Coders, which is subscribed to GNU Mailman. https://bugs.launchpad.net/bugs/1184376 Title: REST server crashes on "reopen" To manage notifications about this bug go to: https://bugs.launchpad.net/mailman/+bug/1184376/+subscriptions

2 10

[Bug 1181498] [NEW] subject_prefix is not part of the IMailingList interface
by Aurélien Bompard April 22, 2014

April 22, 2014

Public bug reported: In mailman/interfaces/mailinglist.py, in the IMailingList interface, there's a display_name property whoose documentation reads: "This is used in locations such as the message footers and Subject prefix.". It's actually not true, in the MailingList model class (in mailman/model/mailinglist.py) there's a subject_prefix attribute for this value, and it's the one that Postorius edits. Thus the interface is incomplete (which is not a big problem) and the documentation is misleading (which is a bigger problem IMHO). ** Affects: mailman Importance: Undecided Status: New ** Tags: documentation mailman3 ** Tags added: documentation 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/1181498 Title: subject_prefix is not part of the IMailingList interface To manage notifications about this bug go to: https://bugs.launchpad.net/mailman/+bug/1181498/+subscriptions

2 2