Postfix.py and aliases (#CREATED:xxx)

[Eawf]

CentOS4.2 Installed Mailman2.1.5 via YUM.
Postfix
Configured Automatic aliases file creation.

When creating list from ./newlist, no problems.
When creating list from web, got bug page, and
discovered an alias entry, "#CREATED:...".

While not being a Python programmer, glancing through
the Postfix.py file, it looks like when looping through
the STANZA section to create the aliases file, the
commented #CREATED: line isn't ignored, causing the
record to be created.

[http://sourceforge.net/tracker/index.php?func=detail&aid=1459022&group_id=103&atid=100103]

[Mark Sapiro (msapiro)]

Both bin/newlist and the Cgi create script invoke the exact
same Mailman/MTA/Postfix.py module to update the aliases.
The only difference is the user/group under which it runs.
Thus, this is most likely a problem with the 'mailman' group
not having write permission on data/aliases*. Both
data/aliases and data/aliases.db should be group owned by
the 'mailman' group and be group writable.

This may be FAQ 6.9 -
http://www.python.org/cgi-bin/faqw-mm.py?req=show&file=faq06.009.htp

If this is not a permissions issue, please post the
traceback from the "bug" from Mailman's error log.

Note that the "#CREATED:" line in the aliases file is a
comment and is intended to be there.

[Eawf (eawf)]

Interestingly, when I commented out both of the "CREATED:
print statements, the problem went away.

Only thing I can figure from that is that programmatically,
when creating the aliases file, the commented CREATED: line
is not being ignored or skipped over. Again, glancing
through the code I see checking for the STANZA lines, but
not the CREATED line.

[Mark Sapiro (msapiro)]

I think the line you're referring to is

    print >> fp, '# CREATED:', time.ctime(time.time())

Which occurs in two places, one for the 'aliases' file and
one for the 'virtual-mailman' file if any. It writes a line like

# CREATED: Wed Mar 29 19:31:24 2006

(with the current date and time) into the file. This line in
the 'aliases' or 'virtual-mailman' file begins with '#' and
thus is a comment in that file just as are the lines

# STANZA START: listname

preceding the # CREATED line and

# STANZA END: listname

following the actual data. Why is this "# CREATED: ..." line
a problem?

Going back to your original issue, what is the traceback
from the "bug" from Mailman's 'error' log?

I don't think this 'bug' has anything to do with the print
statements you commented or with the presence or absence of
the "# CREATED: ..." line in the resultant file.

The "# CREATED: ..." line is documentation. The code that
looks at the "# STANZA: ..." lines is figuring out what to
delete from the files when you delete a list.