Mailman 3 December 2006 - Mailman-Developers

Pickles begone
by Barry Warsaw 05 Jan '07

05 Jan '07

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 I'm about to merge my SQLAlchemy branch to the trunk. I'm happy enough with where this is going to commit to this approach going forward. I wrote about it a bit in a news article on wiki.list.org and I'll try to elaborate there when things settle down. This stuff could definitely use more testing, but the web u/i and sending mail through the system work for me. If you're using the trunk regularly now, you will notice that there's no migration path yet for moving from the pickle-based storage to SQLAlchemy based storage. This is what the import/export scripts are for. You notice though that only bin/export is written yet. bin/import is tops on my list after the merge. I've also committed bin/export.py to the Mailman 2.1 branch. This will be how you migrate MM2.1 lists to the MM trunk. Our next release of MM2.1 will have this, so my plan for migration will be to install the latest 2.1.x (x > 9), run bin/export.py, then feed the resulting XML to the trunk's bin/import. For now, you can start from scratch with bin/newlist or creating a list through the web. From there the trunk should look and act just like you're used to. As always, let us know what you find. I'm eager to fix problems quickly. Please don't use the SF bug tracker for problems you encounter on the trunk. It would also be interesting to know if you try a SQLAlchemy backend other than SQLite. - -Barry -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.5 (Darwin) iQCVAwUBRZWUFHEjvBPtnXfVAQIEJwP/WkB2jY3Ww2YR96oDejC/ZSSWwbct8ENb k/kpcoYm9V2ra36WUiHt0pfFLyhMk0pRXtBh/oScooLiZJu9AA2wnD046I5AJZLt 5DS+YJI2B5c4pn4pO80w5JeAUgQfiYIhafs6u/Oy3fDzag6Rp698Dt/Bv62V8Mly hteE9Iu8xcs= =plDI -----END PGP SIGNATURE-----

5 12

fixing news threading bug
by Daniel Poelzleithner 22 Dec '06

22 Dec '06

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi, regarding: https://sourceforge.net/tracker/?func=detail&aid=1256272&group_id=103&atid=… What do the main mailman devs think about it ? I would write a patch, is it ok to use sqlite to store the table, since its already part of python2.5 and usually installed everywhere ? kindly regards Daniel -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.5 (GNU/Linux) Comment: GnuPT 2.7.2 Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFFjBOXy/mkIQp7AD0RAlDlAKCtXO79UEM4yPGPFrYlr5Fn70hQoACghCXa c8IjXGi0oh6uh8Fx0Dg0i2A= =JuEA -----END PGP SIGNATURE-----

1 0

New Internet Casino page. $10 free on signup!!!!!
by estlibit＠quickstockpower.com 22 Dec '06

22 Dec '06

1 0

Multiple instances of mailman : some thoughts for improving the current status.
by Clement Hermann 18 Dec '06

18 Dec '06

Hello, I need multiple instances of mailman for separate name space between virtual domains, as I'm not willing to patch if I can avoid it (I'd rather migrate to a virtual-hosting capable mailman once it goes out). However, installing multiple mailman instances from source with a different $PREFIX could make things difficult in a production environment, especially regarding security updates. In a perfect world (from the sysadmin point of vue) the prefix should be configurable through an environment variable or a configuration file (but the latter would be more complicated, as you would have to be able to choose the mm_cfg.py from the command line). That way, it would be trivial to set up several instances by copiying data files and linking binaries / script. a simple wrapper (shell, or C, or whatever) could then call the right binaries from the current path or an additional parameter. A security upgrade would then be just a matter of reinstalling the files in the master installation I'm far from being a python coder (I know perl and shell a lot better), but I can't see from the code where that could cause a problem. The C CGI won't represent a problem either, just need to use the environment with perhaps a default value for the $prefix value instead of hard-coding it at compilation time. Is there anything preventing us to modify the code in such a way ? Am I missing something obvious, or underestimating the difficulties ? Please tell me what you think. Cheers, -- Clément Hermann (nodens)

1 0

MySQLMemberships.py
by Fil 15 Dec '06

15 Dec '06

Hello, I have upgraded the MysqlMemberships.py member adaptor, so that it uses only one connection for all lists, and not one connection per list (as it did before). Everything works fine (only one day of operation though), except when the mysql link is broken (I stopped mysqld to check that). Then I get this error: self._prodServerConnection() File "/var/local/mailman/Mailman/MysqlMemberships.py", line 132, in _prodServerConnection mm_cfg.connexion.close() _mysql_exceptions . ProgrammingError : closing a closed connection I guess I could use help from someone who knows more python than me. If you want to have a look at the code it's here: http://trac.rezo.net/trac/rezo/browser/Mailman/MySQLMemberAdaptor/MysqlMemb… -- Fil

1 1

Re: [Mailman-Developers] [Mailman-Users] Mailman archive messages(not rm, but install!)
by Barry Warsaw 08 Dec '06

08 Dec '06

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Dec 8, 2006, at 11:10 AM, Paul Tomblin wrote: >> In that case, shouldn't any message that reaches mailman with an >> unescaped From_ line in the body already be handled properly? It >> seems like something else must be borked. That or all of the >> messages >> in a list mbox that contain unescaped From_ lines got there from >> really old versions of Mailman/python. > > That is distinctly possible. The archives in question go back to > 1998. I > didn't keep track of when the *last* unescaped From_ line was put > in the > archives. Sorry, I should have been clearer that the /default/ behavior of the generator is to mangle From_ lines. So it's true that nothing in Mailman should need to be changed. However, it's also true that in the distant past, there were some bugs in the mbox implementation which would cause broken mbox files to be written. A quick scan through the svn logs jogs my memory: r6341 on 2003-04-17 was added to fix a message separation bug. I don't know how long that bug was lurking, but the fix puts it just before the 2.1.2 release according to the NEWS file. I'll bet that it existed from 2.1 final (Dec 2002) until 2.1.2 (Apr 2003), the latter which was probably released specifically to fix this problem! Note that this bug had no effect on the archiving of new messages on the fly. Those always got archived correctly. But the message was appended to the mbox file incorrectly which meant that if you regenerated your archives, you'd be screwed. This was what bin/ cleanarch was intended to fix. BTW, one less ambitious way to participate here to help fix things would be to improve bin/cleanarch. At the very least, you should be able to run that script and get an mbox file that bin/arch can use to DTRT. It would also be nice if bin/arch was able to compensate for running out of memory, possibly by changing it to fork a sub-process to do the actual archiving with the parent process pre-chunking the workload for the child. Anyway, I'm cc'ing mailman-developers. Further discussion of how to improve matters should be conducted on that list (and mailman-users should be removed). - -Barry -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.5 (Darwin) iQCVAwUBRXmlMHEjvBPtnXfVAQIFYAP/W2LEOrKhqrB6sDniHKADAV5iMuLm19zu nUkvrJpOumD78+tRDa1DCQG8RaCSAZ7bNkTA2VwIUgcX1I4+9d7ylklonQSiRJzB xbg+OBD5+x5q+Cdo9qX1dhlGWTdmrSReN0CLRx6408JX8qtXhIh+3S0f3tG44bYE lB76OX4HPXo= =nhI8 -----END PGP SIGNATURE-----

1 0

Search engine + authentication
by Kiss Gabor (Bitman) 06 Dec '06

06 Dec '06

Dear folks, I've just combined Namazu (http://www.namazu.org/index.html.en) with Mailman as described in Tom Morrison's article (http://mail.python.org/pipermail/mailman-users/2004-June/037584.html) However I found that unathenticated/unauthorized users can search in any list archive. After a few experimental solutions finally I rewrote Lindsay Haisley's nmz_wrapper.cgi script (http://www.fmp.com/namazu/). Now it is called /usr/lib/mailman/Mailman/Cgi/search.py and uses the same authentication mechanism as /usr/lib/mailman/Mailman/Cgi/private.py does. I use Debian binary version of mailman and I did not want to recompile the package. However I needed a new cgi-wrapper named /usr/lib/cgi-bin/mailman/search to call search.py with proper setgid settings. So I just patched an other wrapper: perl -p -e 's/rmlist/search/g' rmlist > search (Yes I know this is a rude hack. :-) My question is: is it possible to add such a wrapper to the next release as hook for any third party search engine? Gabor

1 0

Postfix/virtual-mailman not compatible when using dedicated destination (like "localhost")
by Joe Peterson 05 Dec '06

05 Dec '06

I wanted to post this to the developer list and see if there have been any previous thoughts on my situation and also if there are related features being developed for 3.0 (if so, there's probably no point in patching 2.x). Anyway, my postfix configuration, although probably not rare, seems to be different that many. In my case, all of my domains are "virtual". I have real unix users, and in order to have the same email account name in different domains, I make a user with "name-domain". In my postfix virtual file, I map "name@domain" to name-domain, however, since I do not want to pick one of my actual domains as the default destination, I set $mydestination to "localhost", requiring me to map "name@domain" to "name-domain@localhost". The benefit is that there is no domain to which "name-domain@domain" would work (I do not want this to work as a side effect). Sorry for the long description!! :) Anyway, give the above, without modifying MTA/Postfix.py (and it's a trivial mod, I know), there is no way to make Mailman's automatic postfix alias stuff completely work. I now have my own auto-generated virtual file and I do not use virtual-mailman, but I'd rather take advantage of it if I could. There's another guy in my situation who patched Postfix.py, and I recently contacted him to see if he's every communicated with you guys or submitted a patch. He said no, but he's be glad if I did. So that brings me here! :) I looked at the latest svn code, and I did not see something like this implemented. His site, which describes all of this, is: http://www.qnan.org/~pmw/howtos/mailman-virtuals.html I've submitted patches for Mailman before, and I'd be glad to do one again, but I wanted to ask: 1) Do you see benefit to adding a variable that would define this right-hand-side domain (e.g. "localhost") for virtual-mailman? 2) What is a good name for the variable? (POSTFIX_VIRTUAL_DESTINATION?) Thanks, Joe

3 8

The age old question - virtual domains
by Michael Kabot 02 Dec '06

02 Dec '06

Hi, I've searched quite a bit and found that virtual domains is still not supported in the current version of Mailmain (2.1.9). It sounds like it is around the corner in v2.2, whenever that is. I've also tried to use some of the "patches" that others have pointed to without success. I have a short term need for virtual domains in Mailman and believe I have a quick fix that I could use some brief expertise in implementing, thus why I am emailing this group. Here is my configuration - Dedicated Servers at GoDaddy.com - Fedora Core 4 (2.6.17 kernel) - Mailman 2.1.5 - Python 2.4.3 - Plesk 8.1 - Qmail 1.03 Here is my idea - create all mailman lists in the format <virtual domain name>-<list name> - This can be done right through Plesk - setup a mail alias for the virtual domain with the name <listname> that forwards to the <virtual domain name>-<list name>@<virtual domain name> email address - Again, this can be done right through Plesk - Setup a regexp for the mailman list that looks for the <virtual domain name>-<list name> format - Presto, virtual domains working in mailman ! I will not be using ANY of the email admin functions, i.e. -request, -subscribe, -unsubscribe, etc.. I've got this all working well. The ONLY issue that I can see is that I need to have mailman munge two email addresses... - In the mail that is forwarded to the list members, the Reply-To and "on behalf of" of the person who sent the email needs to be put back in. Based on the forwarding, it is showing up as <virtual domain name>-<list name> - In that same email, change the <virtual domain name>-<list name>-bounces email to just <list name>-bounces. Can someone help me with two things 1. I need a pointer to where in the code to do the munge above. 2. Any gotcha's you can think of to this idea? Thanks in advance, Michael Kabot President - SOAR Scouting Online Affordable & Reliable mkabot(a)soarol.com 585-388-0211 www.soarol.com <http://www.soarol.com/>

2 1