Mailman 3 March 2002 - Mailman-Developers

Thread not shown allthough In-Reply-To present
by Bram de Kruijff 07 Mar '02

07 Mar '02

Hi all, allthough I found a lot of questions about why mailman/pipermail doesn't show threads when mail is transferred by outlook/exchange in the archives, but unfortuanatly no answer that solves the problem ;( I read in one of the posts that threading is done based on; - Message-ID - References - In-Reply-To But then why doesn't it work? According to the mbox of my archive messages sent by outlook/exchange users don't contain a References header but DO contain Message-ID and In-Reply-To headers. In the same archive messages sent with mutt do contain a References header and are shown threaded Does pipermail realy check the In-Reply-To? Is something wrong with the format used? eg; In-Reply-To: <6C9EC88CAD726B49A43EE2E2B2AD6E5383E93C(a)maul.gx.nl> any ideas? thanks in advance! Bram _____________________________________________________________ Bram de Kruijff___________ <GX> creative online development _______________________ http://www.gx.nl | mailto:bramk@gx.nl

1 0

list_members doesn't return the whole membership list
by Marc MERLIN 06 Mar '02

06 Mar '02

root@gandalf:/var/local/mailman/bin# ./dumpdb ../lists/test3/config.pck | grep user_options 'user_options': {'marc(a)merlins.org': 266}, root@gandalf:/var/local/mailman/bin# ./list_members test3 root@gandalf:/var/local/mailman/bin# On bigger lists, list_members returns some of the list membership. It's not just the digest or non digest members, it looks kinda random. I need to get some sleep, so I'll leave this for someone else to debug and fix Marc -- Microsoft is to operating systems & security .... .... what McDonalds is to gourmet cooking Home page: http://marc.merlins.org/ | Finger marc_f(a)merlins.org for PGP key

2 3

Python-2.2 disaster
by Carson Gaspar 06 Mar '02

06 Mar '02

<rant> What !@#$% idiot did the python 2.2 build system!? It bombs out trying to compile the extensions, because _someone_ got "clever" and decided to write the build system in Python. And got it wrong. So it doesn't compile or link properly. And after chasing through _8_ levels of indirection without figuring out where it tries adding library arguments, I just give up! For those who care, it's going insane trying to build _socket, and _curses, and probably something else. _socket is trying to link against openssl, but my openssl libraries aren't where it expects them to be, and if I hand-edit setup.py (yes, hand-edit - ARRRRGH!) to put the correct paths in, it still doesn't work, I suspect because it's not using -R or setting LD_RUN_PATH. I haven't even looked at the other errors yet. </rant> So, just how "recommended" is python 2.2? Can I just use my python 2.0 installation until someone else makes this nightmare go away? -- Carson Gaspar - carson(a)taltos.org Queen Trapped in a Butch Body

8 16

Small patch
by mss＠mawhrin.net 06 Mar '02

06 Mar '02

There are two same questions, but in one place there is a space at the end, while in the other there is not. I propose to remove the space in the first case. -- Misha Index: admin.py =================================================================== RCS file: /cvsroot/mailman/mailman/Mailman/Cgi/admin.py,v retrieving revision 2.60 diff -u -b -r2.60 admin.py --- admin.py 5 Mar 2002 04:57:51 -0000 2.60 +++ admin.py 6 Mar 2002 17:45:50 -0000 @@ -1053,7 +1053,7 @@ ]) table.AddRow([ # td 1 - Label(_('Send notifications to the list owner? ')), + Label(_('Send notifications to the list owner?')), # td 2 RadioButton('send_notifications_to_list_owner', 0, not mlist.admin_notify_mchanges).Format()

1 0

updated patch: remove member from all the lists
by Marc MERLIN 06 Mar '02

06 Mar '02

This version of the patch add --fromall which lets you remove a specified Email from all your lists (replaces my _alllists_ special hack) --- ../../src/mailman-cvs/bin/remove_members Fri Sep 7 16:18:47 2001 +++ remove_members Mon Mar 4 23:02:17 2002 @@ -19,7 +19,7 @@ """Remove members from a list. Usage: - remove_members [options] listname [addr1 ...] + remove_members [options] ([-a] listname|--fromall) [addr1 ...] Options: @@ -31,6 +31,12 @@ --all -a Remove all members of the mailing list. + (mutually exclusive with --fromall) + + --fromall + Removes the given addresses from all the lists on this system + regardless of virtual domains if you have any. + (mutually exclusive with -a/--all) --help -h @@ -47,6 +53,7 @@ import paths from Mailman import MailList +from Mailman import Utils from Mailman import Errors from Mailman.i18n import _ @@ -77,17 +84,16 @@ def main(): try: opts, args = getopt.getopt( - sys.argv[1:], 'af:h', ['all', 'file=', 'help']) + sys.argv[1:], 'af:h', ['all', 'fromall', 'file=', 'help']) except getopt.error, msg: usage(1, msg) if len(args) < 1: usage(1) - listname = args[0].lower().strip() - addresses = args[1:] filename = None all = 0 + alllists = 0 for opt, arg in opts: if opt in ('-h', '--help'): @@ -96,33 +102,55 @@ filename = arg elif opt in ('-a', '--all'): all = 1 - + elif opt == '--fromall': + alllists = 1 + + # You probably don't want to delete all the users of all the lists -- Marc + if all == 1 and alllists == 1: + usage(1) + + if alllists: + addresses = args + else: + listname = args[0].lower().strip() + addresses = args[1:] + + if alllists: + listnames = Utils.list_names() + else: + listnames = [ listname ] + if filename: try: addresses = addresses + ReadFile(filename) except IOError: print _('Could not open file for reading: %(filename)s.') - try: - # open locked - mlist = MailList.MailList(listname) - except Errors.MMListError, e: - print _('No such list: %(listname)s') - sys.exit(1) - - if all: - addresses = mlist.getMembers() - - try: - for addr in addresses: - try: - mlist.ApprovedDeleteMember(addr) - except Errors.MMNoSuchUserError: - print _("User `%(addr)s' not found.") - finally: - # Hmm, should it be all or nothing? - mlist.Save() - mlist.Unlock() + for listname in listnames: + try: + # open locked + mlist = MailList.MailList(listname) + except Errors.MMListError: + print _('Error opening list %(listname)s... skipping.') + continue + + if all: + addresses = mlist.getMembers() + + try: + for addr in addresses: + try: + mlist.ApprovedDeleteMember(addr) + except Errors.MMNoSuchUserError: + if not alllists: + print _("User `%(addr)s' not found.") + else: + if alllists: + print _("User `%(addr)s' removed from list '%(listname)s'.") + finally: + # Hmm, should it be all or nothing? + mlist.Save() + mlist.Unlock() -- Microsoft is to operating systems & security .... .... what McDonalds is to gourmet cooking Home page: http://marc.merlins.org/ | Finger marc_f(a)merlins.org for PGP key

2 1

buglet, changing options on an already deleted member
by Marc MERLIN 06 Mar '02

06 Mar '02

I bring up an admin page, from the shell, I remove a user I change the user options in the web form, and click submit Traceback (most recent call last): File "/var/local/mailman/scripts/driver", line 82, in run_main main() File "/var/local/mailman/Mailman/Cgi/admin.py", line 166, in main change_options(mlist, category, subcat, cgidata, doc) File "/var/local/mailman/Mailman/Cgi/admin.py", line 1318, in change_options mlist.setMemberName(user, newname) File "/var/local/mailman/Mailman/OldStyleMemberships.py", line 315, in setMemberName self.__assertIsMember(member) File "/var/local/mailman/Mailman/OldStyleMemberships.py", line 113, in __assertIsMember raise Errors.NotAMemberError, member NotAMemberError: marc(a)merlins.org The error is correct, but mailman should probably return a simple error in the web page instead of a traceback. Marc -- Microsoft is to operating systems & security .... .... what McDonalds is to gourmet cooking Home page: http://marc.merlins.org/ | Finger marc_f(a)merlins.org for PGP key

2 1

How do I turn off line wrapping on template files?
by Carson Gaspar 06 Mar '02

06 Mar '02

(Yesterday's CVS, Python 2.1.2, Solaris 8 SPARC) I'm trying to customize the e-mail messages that go out in response to a subscription (verify.txt and subscribeack.txt). Out-of-the-box, they get mangled by Util.wrap(), causing havoc with our attempts at formatting. For example, we can't have: Please contact one of the following people for more information: bozo(a)hotmail.com jesus(a)yahoo.com god(a)heaven.com It all appears on one line. Looking at the code, it looks like my choices are to wrap or not wrap - there's no magic code for "No, I really want a CRLF here!" That being the case, I tried adding raw=1 to the maketext calls for verify.txt (in MailList.py) and for subscribeack.txt (in Deliverer.py). It worked for verify.txt (huzzah!), but is still wrapping subscribeack.txt (wah!). So, am I missing some cool feature of wrap? If not, is there some reason adding raw=1 isn't working? -- Carson

2 4

Re: [Mailman-Developers] list_members doesn't return the whole membership list
by Dan Mick 05 Mar '02

05 Mar '02

Just sent this to Barry early Monday morning: Date: Mon, 4 Mar 2002 09:58:20 -0800 (PST) From: Dan Mick <dmick@utopia74> Subject: bug: list_members has a backwards test To: barry(a)wooz.org The ambiguously-named 'statusp' returns 'true' if the address should be filtered, so the tests are backwards: Index: list_members =================================================================== RCS file: /cvsroot/mailman/mailman/bin/list_members,v retrieving revision 2.4 diff -u -r2.4 list_members --- list_members 23 Feb 2002 06:40:44 -0000 2.4 +++ list_members 4 Mar 2002 17:58:09 -0000 @@ -179,14 +179,14 @@ rmembers.sort() for addr in rmembers: # Filter out nomails - if not statusp(mlist, addr, why): + if statusp(mlist, addr, why): continue print >> fp, addr if digest: dmembers.sort() for addr in dmembers: # Filter out nomails - if not statusp(mlist, addr, why): + if statusp(mlist, addr, why): continue # Filter out digest kinds if mlist.getMemberOption(addr, mm_cfg.DisableMime): > root@gandalf:/var/local/mailman/bin# ./dumpdb ../lists/test3/config.pck | grep user_options > 'user_options': {'marc(a)merlins.org': 266}, > root@gandalf:/var/local/mailman/bin# ./list_members test3 > root@gandalf:/var/local/mailman/bin# > > On bigger lists, list_members returns some of the list membership. It's not just > the digest or non digest members, it looks kinda random. > > I need to get some sleep, so I'll leave this for someone else to debug and > fix

1 0

Re: [Mailman-Developers] Python-2.2 disaster
by Dan Mick 05 Mar '02

05 Mar '02

> FYI, 2.2 is completely DOA on Solaris 8/SPARC. I'm using 2.2 daily on Solaris 8/SPARC.

1 0

Current CVS update error
by Carson Gaspar 05 Mar '02

05 Mar '02

It looks like the install / update procedure does not create qfiles/in, so I get the following when I run bin/update: updating old qfiles Traceback (most recent call last): File "/tools/mailman-cvs-20020304/bin/update", line 533, in ? errors = main() File "/tools/mailman-cvs-20020304/bin/update", line 464, in main update_qfiles() File "/tools/mailman-cvs-20020304/bin/update", line 380, in update_qfiles os.rename(oldmsgfile, newmsgfile) OSError: [Errno 2] No such file or directory After I manually create qfiles/in, it runs correctly. Also, check_perms does _not_ complain about qfiles/in being missing. -- Carson Gaspar - carson(a)taltos.org Queen Trapped in a Butch Body

2 2