Mailman 3 July 1998 - Mailman-Users

Multiple copies of messages?
by William Bradford 02 Aug '98

02 Aug '98

I'm running 1.0b5 with apache 1.3 and python 1.5.1 on a DEC UNIX 4.0D machine (AlphaStation 200 166mhz), and some of my list members are getting multiple copies of the same message.. Also, I found that you cant "en masse" add more than 25 or so people at a time, or the operation bombs out with an "operation would block" message on the error bug page returned by the web browser... bill -- Bill Bradford * mrbill(a)frenzy.com ------------------------------------------- "One World, One Web, One Program" - Microsoft Promotional Ad "Ein Volk, Ein Reich, Ein Fuhrer" - Adolf Hitler

2 5

Reposting Form Data
by Barry A. Warsaw 01 Aug '98

01 Aug '98

Maybe someone out there with more HTTP experience than me can answer this question. If I have a CGI script that issues a redirect using the `Location:' header, I can get the Web browser to go to the relocated page with no problem. It even still shows the original URL in browser. Cool so far. However, if what I'm doing is posting some form data, the browser gets redirected, but the form data doesn't get reposted to the new URL. Verified under NS4/Unix and IE3. This is a real pain for something I'm trying to add to Mailman. I'm sure there's some haughty security justification for the non-reposting of form data, so the natural question is: can I get around this? :-) Thanks, -Barry

2 1

problem with "Edit Options"
by William Bradford 01 Aug '98

01 Aug '98

Tried to do the "Edit Options" section, and edit my personal user settings on a test list, and I got this output: (http://area51.mrbill.net/mailman/subscribe/test) We're sorry, we hit a bug! If you would like to help us identify the problem, please email a copy of this page to the webmaster for this site with a description of what happened. Thanks! Traceback: Traceback (innermost last): File "/usr/local/mailman/scripts/driver", line 97, in run_main main() File "../Mailman/Cgi/subscribe.py", line 95, in main call_script('options', [list._internal_name, member]) File "../Mailman/Cgi/subscribe.py", line 65, in call_script list.Unlock() AttributeError: Unlock Environment variables: Variable Value DOCUMENT_ROOT /usr/local/apache/share/htdocs CONTENT_LENGTH 49 CONTENT_TYPE application/x-www-form-urlencoded PATH_TRANSLATED /usr/local/apache/share/htdocs/test/ REMOTE_ADDR 198.214.63.63 SERVER_SOFTWARE Apache/1.3.1 (Unix) GATEWAY_INTERFACE CGI/1.1 HTTP_ACCEPT_LANGUAGE en REMOTE_PORT 17021 SERVER_PORT 80 HTTP_CONNECTION Keep-Alive HTTP_USER_AGENT Mozilla/4.05 [en] (X11; I; Linux 2.0.33 i586) HTTP_ACCEPT_CHARSET iso-8859-1,*,utf-8 HTTP_ACCEPT image/gif, image/x-xbitmap, image/jpeg, image/pjpeg, image/png, */* REQUEST_URI /mailman/subscribe/test/ PATH /usr/local/bin:/bin:/sbin:/usr/sbin:/usr/bin QUERY_STRING SERVER_PROTOCOL HTTP/1.0 PATH_INFO test/mrbill(a)frenzy.com HTTP_HOST area51.mrbill.net REQUEST_METHOD POST SCRIPT_NAME /mailman/subscribe SERVER_ADMIN root(a)area51.mrbill.net SCRIPT_FILENAME /usr/local/mailman/cgi-bin/subscribe PYTHONPATH /usr/local/mailman HTTP_REFERER http://area51.mrbill.net/mailman/listinfo/test SERVER_NAME area51.mrbill.net -- Bill Bradford * mrbill(a)frenzy.com ------------------------------------------- "One World, One Web, One Program" - Microsoft Promotional Ad "Ein Volk, Ein Reich, Ein Fuhrer" - Adolf Hitler

2 1

Potential gotcha
by Barry A. Warsaw 31 Jul '98

31 Jul '98

Corbett has been very kind in allowing me access to his machine to debug his problems. Since others have reported similar problems, I want to give you a heads up. Please double check that the programs in /home/mailman/cgi-bin are group owned by `mailman' and have the group sticky bit set. This is imperative; Mailman just won't work without this. I know need to figure out why the install procedure is broken on some systems, and how the driver script can better report and recover from these problems. Thanks Corbett! -Barry

1 1

Re: [Mailman-Users] Multiple copies of messages?
by Milan Zamazal 31 Jul '98

31 Jul '98

>>>>> "BAW" == Barry A Warsaw <bwarsaw(a)CNRI.Reston.Va.US> píše: BAW> We know that 1.0b5 has a bug that subscription acknowledgements BAW> are sent out twice. We haven't tracked this problem down yet. BAW> Are you saying that your members are also getting some posts BAW> more than once? I haven't seen that on my test lists, but it BAW> could still be happening (and possibly caused by the same bug BAW> that's sending out subscript acks twice). I had the same problem (some people received even about 10 copies of the subscription message), and yes, some members received also messages twice. The latter problem disappeared after I set number of outgoing connections to 1 instead of default 5 in Mailman. However mass subscriptions are still sent more than once to some users. Milan Zamazal

1 0

gid error
by m* 31 Jul '98

31 Jul '98

i keep getting this error. i thought i had followed the dirs but i must be missing something: Jul 30 11:58:47 orifice Mailman mail-wrapper: Failure to exec script. WANTED gid 8, GOT gid 1. (Reconfigure to take 1? i let the configure script determine the gid and it got mail's gid 8 but it appears that the script is trying to run as daemon's gid. i configured mailman once before ( i believe it was an earlier version ) but did not have this problem. thanks so much, mark

2 1

Re: [Mailman-Users] Mailman problem
by Barry A. Warsaw 31 Jul '98

31 Jul '98

>>>>> "L" == Luke <luke(a)indy.net> writes: L> none of these failed Sorry Luke, I've been out sick since just about after our last email exchange. Are you still having problems? If you're getting that low-level bug report then there *must* either be problems importing sys, or something that's causing the functions in the except clause of run_main() to fail. I still suspect that logger or multi are not getting instantiated correctly. Please try this version of the driver script and see if that gives you any more clues as to what's up. This will print error information to whatever stderr is for your CGI scripts. Let me know what you find out. -Barry #! /usr/bin/env python # # Copyright (C) 1998 by the Free Software Foundation, Inc. # # This program is free software; you can redistribute it and/or # modify it under the terms of the GNU General Public License # as published by the Free Software Foundation; either version 2 # of the License, or (at your option) any later version. # # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU General Public License for more details. # # You should have received a copy of the GNU General Public License # along with this program; if not, write to the Free Software # Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. # Useful for debugging. When an error occurs, this attaches the file name to # the exception string and re-raises (using the bogus Python 1.5 semantics) # this may be unnecessary in Python 1.5.2 ## realopen = open ## def open(filename, mode='r', bufsize=-1, realopen=realopen): ## try: ## return realopen(filename, mode, bufsize) ## except IOError, e: ## strerror = e.strerror + ': ' + filename ## e.strerror = strerror ## e.args = (e.args[0], strerror) ## # Python 1.5 ## import sys ## raise e, None, sys.exc_info()[2] ## # Python 1.5.1 ## #raise ## import __builtin__ ## __builtin__.__dict__['open'] = open # This standard driver script is used to run CGI programs, wrapped in code # that catches errors, and displays them as HTML. This guarantees that # (almost) any problems in the Mailman software doesn't result in a Web server # error. It is much more helpful to generate and show a traceback, which the # user could send to the administrator, than to display a server error and # have to trudge through server logs. # Note: this isn't 100% perfect! Here are some things that can go wrong that # are not caught and reported as traceback-containing HTML: # # - This file could contain a syntax error. In that case, you would indeed # get a Web server error since this file wouldn't even compile, and there's # no way to catch that. # # - The sys module could be royally screwed. Either we couldn't import it, or # it didn't have a sys.stderr attribute. Both those would indicate serious # problems in the Python installation. These won't generate Web server # errors, but neither will they give meaningful tracebacks. # # - We couldn't import the traceback module, or traceback.print_exc() # failed. Same diagnosis and effect as with sys being broken. # # I consider all these pretty unlikely. Note that it is also possible that # the environment variables wouldn't be printed, perhaps because os couldn't # be imported or there was some problem with os.environ. Again, not likely. def run_main(): try: # These will ensure that even if something between now and the # creation of the real logger below fails, we can still get # *something* meaningful logger = sys.stderr multi = sys.stderr # insert the relative path to the parent of the Mailman package # directory, so we can pick up the Utils module import os # sys gets imported at module level below sys.path.insert(0, os.pardir) # map stderr to a logger, if possible from Mailman.Logging.StampedLogger import StampedLogger from Mailman.Logging.MultiLogger import MultiLogger logger = StampedLogger('error', label='admin', manual_reprime=1, nofail=0) multi = MultiLogger(sys.__stdout__, logger) # The name of the module to run is passed in argv[1]. What we # actually do is import the module named by argv[1] that lives in the # Mailman.Cgi package. That module must have a main() function, which # we dig out and call. # scriptname = sys.argv[1] # See the reference manual for why we have to do things this way. # Note that importing should have no side-effects! pkg = __import__('Mailman.Cgi', globals(), locals(), [scriptname]) module = getattr(pkg, scriptname) main = getattr(module, 'main') try: main() except SystemExit: # this is a valid way for the function to exit pass except: print_traceback(logger, multi) print_environment(logger) def print_traceback(logger, multi): print """\ Content-type: text/html <p><h3>We're sorry, we hit a bug!</h3> <p>If you would like to help us identify the problem, please email a copy of this page to the webmaster for this site with a description of what happened. Thanks! <h4>Traceback:</h4> <p><pre> """ logger.write('[----- Traceback ------]\n') try: import traceback # in normal situation, this will get logged to the MultiLogger created # above, which will write the data to both the real live stdout, and # the StampedLogger traceback.print_exc(file=multi) except: multi.write('[failed to get a traceback]\n') print '\n\n</pre>' def print_environment(logger): try: import os print '''\ <p><hr><h4>Environment variables:</h4> <p><table> <tr><td><strong>Variable</strong></td> <td><strong>Value</strong></td></tr> ''' logger.write('[----- Environment Variables -----]\n') for varname, value in os.environ.items(): print '<tr><td>', varname, '</td><td>', value, '</td></tr>' logger.write('\t%s: %s\n' % (varname, value)) print '</table>' except: print '<p><hr>[environment variables are not available]' try: import sys try: run_main() finally: # this is probably not strictly necessary since the script is exiting # soon anyway sys.stderr = sys.__stderr__ except: # Jeez, we couldn't even import sys, or sys didn't have a stderr # attribute! print """\ Content-type: text/html <p><h3>We're sorry, we hit a bug!</h3> Mailman experienced a very low level failure and could not even generate a useful traceback. Please report this to the Mailman administrator at this site. """

1 0

None
by Oliver Thuns 30 Jul '98

30 Jul '98

confirm 117683

1 0

Mailman problem
by Luke 29 Jul '98

29 Jul '98

When trying to do any of the CGI stuff in 1.0b4 i get the following. I am able to view all the "top level" CGI pages, but i get this when submitting changes in /admin/ now as well as in /listinfo/: We're sorry, we hit a bug! If you would like to help us identify the problem, please email a copy of this page to the webmaster for this site with a description of what happened. Thanks! Traceback: Content-type: text/html We're sorry, we hit a bug! Mailman experienced a very low level failure and could not even generate a useful traceback. Please report this to the Mailman administrator at this site. Luke luke(a)indy.net "Celebrate we will because life is short but sweet for certain"---Two Step

2 4

problems
by Luke 28 Jul '98

28 Jul '98

sorry. . .disregard the last message. . .hit cntrl-x rather than cntrl-c :) i'll post again when i get the error to rear its ugly head again :P Luke luke(a)indy.net "Celebrate we will because life is short but sweet for certain"---Two Step

1 0