[Mailman-Users] gzipped archive files.

[Mark Sapiro]

Williams, Paul wrote:

>I noticed that we are gzipping the archive files, nightly
>with
>
>27 3 * * * /usr/bin/python -S /opt/mailman/cron/nightly_gzip
>
>but it also leaves a copy of the uncompressed archives.
>
>Is this the proper behavior?


It is what it is designed to do.


>Is it possible to only have the gzipped archives and not
>the text files?


Yes, but you have to do it outside of Mailman. E.g., set up a cron to
run on say the 3rd day of the month that runs a script that removes
the .txt files for the prior month. There is a small risk to this for
the same reason you can't remove the current month's .txt file. If a
post arrives for a month whose .txt file is missing, it will create a
new .txt file containing only that post, and that file will be gzipped
on the next nightly gzip (or immediately if GZIP_ARCHIVE_TXT_FILES =
Yes in mm_cfg.py), thus losing the prior contents of the .gz file. Of
course, everything is still in the list.mbox/list.mbox file, so it
isn't completely lost, but it's awkward.


>If I comment out the above nightly_gzip line will it
>cause a problem?


No. If there is no .txt.gz file, the archive link will be to the .txt
file. The only reason to have the .txt.gz files is to reduce the size
of the file(s) a user may download, but with many web servers
configured to expand .gz files before serving them, this is not a
consideration.

Of course, the .gz files do save space on the server, but only if the
.txt files are removed which isn't too practical as I note above.

-- 
Mark Sapiro <msapiro at value.net>       The highway is for gamblers,
San Francisco Bay Area, California    better use your sense - B. Dylan