[Kolab-devel] Brain Dump of the Week: One Filter to Rule Them All!

[Sascha Wilde]

Hello List!

it seems the last BDotW was a big success, I got positive feedback and
it spawned some interesting discussions, which is great!  So I decided
to found a new tradition by sending the second one.

After the exiting glance into the future of Kolab Server I reported on
last week, I headed back to (not so) boring ordinary business.  ;-)
This meant primarily looking at various open issues making some small
but worthwhile progress...

The last four days I was unavailable to Kolab, as I was busy with some
other urgent projects (including some real fun hacking, but
unfortunately totally unrelated to Kolab) and next week I will be on
holidays.  So today I reentered the race for the 2.2.3 release and took
full power on the one release critical issue Thomas assigned to me:
Issue973.

The issue is about customization of the Mail Filter's "untrusted from"
messages and it's customer requested stuff[0] so I'm especially happy to
announce that I committed what I think is the final solution to the
server_2_2_branch!  (Please do me a favor and don't look at the date of
the initial message in the issue ... D'oh!)

Well, as you see the issue I worked on is related to my good old friend
Kolab Filter, and it wouldn't be the Kolab Filter if working on one
issue wouldn't raise a bunch of questions, resulting most time a few new
issues.  This time was no difference, so say hello to issue3962,
issue3967 and issue3968, accompanied by a new comment in issue954.

The most interesting issues to work on (though both are reserved for the
next major release, i.e. the CVS HEAD) are 
issue3967: "UNTRUSTED vs. UNAUTHENTICATED, distinction considered harmful"
and issue3968: "Text in Mail Filter Settings misleading".

In 3967 I suggest a basic simplification, which I'd like to see
commented on by those interested in the matter (read: go there and write
a short "Yeah, lets do this!").

3968 is a challenge for the ambitious tech writer: the web admin
interface still fails to give a reasonable description what the Kolab
Filter stuff really does.  Unfortunately the flow diagram showing the
functionality in question consists of no less than 14 boxes, showing
about 19 possible code paths, which is not that easy to put in one or
two short and easily understandable sentences.

Finally, for the code divers who like to get their hands dirty I have a
small treat which I didn't even put into an issue yet:  in the file
Content.php of Kolab Filer
(/kolab/lib/php/Horde/Kolab/Filter/Content.php in OpenPKG based Server)
look out for the line: `if (strpos( $fromhdr, $untrusted )===false) {'
(its line number 449 in the latest version from 2.2-branch).  I have a
very bad feeling about that line as I suspect that MIME encoded data is
compared to a plain string there.  Feel free to send patches or proof me
wrong!  :)

That's all folks!

Today I feel like adding the following DISCLAIMER: 

This report is by no means an objective summary of ongoing affords and
status of the kolab server development, it is just a highly subjective
dump of one developers mind on one arbitrary point of time and subject
to change even in the moment of writing.

cheers
and thanks for contributing to everyone,

sascha

[0] For those of you who always wondered what the `kkc' keyword in the
    issue tracker might mean: it marks issues which are related to
    contracted work for customers.  So solving them helps the successful
    future of Kolab in a special way...
-- 
Sascha Wilde                                          OpenPGP key: 4BB86568
http://www.intevation.de/~wilde/                  http://www.intevation.de/
Intevation GmbH, Neuer Graben 17, 49074 Osnabrück; AG Osnabrück, HR B 18998
Geschäftsführer:   Frank Koormann,  Bernhard Reiter,  Dr. Jan-Oliver Wagner
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://lists.kolab.org/pipermail/devel/attachments/20091125/9637c27f/attachment.sig>