Am 17.07.2014 17:32, schrieb Eric Broch:
Is there a way to call a spam filter in dovecot-lda like maildrop uses the xfilter command within the .mailfilter file to do the same?
are you aware that dspam is dead and the webinterface pratically not maintained and so very likely full of security bugs?
-------- Original-Nachricht -------- Betreff: [Dspam-user] Leaving dspam project Datum: Wed, 09 Jul 2014 16:04:24 +0200 Von: Tom Hendrikx tom@whyscream.net An: dspam-devel@lists.sourceforge.net, dspam-user@lists.sourceforge.net
Hello DSPAM community,
After a few years of being active in the DSPAM community, I've decided to leave the project. Unlike others who silently stopped responding to mails and other inquiries, I'm documenting my reasons in the hopes that it might be of help/interest to others.
When I actively joined the DSPAM project, it was in need of some new blood that helped out on development and other tasks, after most of the original participants of the community takeover from Sensory Networks somehow silently abandoned the project for reasons unknown (to me).
After being active for a few years, I think I've seen most of the problems in the DSPAM project, and tried to address a few without result. The main issues I learned are listed below.
- No active development
There is no one working on the DSPAM code. There are numerous bug reports and feature requests that could be addressed, and some parts in DSPAM have been ripe for refactoring or rewriting for a long time, but nobody is putting any effort into it. There is currently only one developer 'active' within the project, but you rarely see activity from this person. The most code changes in the last 2(?) years are from patches contributed through debian bugreports (thank you, debian maintainer). The fact that no one within the project is actively working on the code, creates a sort of chicken-or-egg problem: without existing activity, no major outside contributions can be expected, since no one is interested in becoming the new sole responsible developer in the project.
- No value seen in making DSPAM more transparent
On several occasions I've raised my voice on making DSPAM more usable to new users, for instance by making error messages more informative, or adding logging of normal operation (i.e. non-errors) so people get a better hang of the product. Changes like this don't 'improve' DSPAM filtering, but the they do make the software easier to use, and its operation more transparent to new and experienced users. Finding issues in software is easier when you actually understand what it's trying to do, which in turn could result in more active users, more useful bug reports, possible more contributed patches and in the end maybe even more developers. Right now DSPAM acts like a sort of black box for anyone not familiar with the source code, and that discourages investigation of odd behavior, which could be any of: configuration errors, bugs or missing (but expected) features.
- No drive on changing the status quo
The situation that DSPAM is in right now, a sort of dormant or zombie (not dead but far from alive) status, is not good. It doesn't attract new users, it doesn't attract new developers, actually it only only supports existing, experienced users. Nobody in the community, both in the project maintainers as in the users community, seems to be interested in changing that. Based on my observations, that seems that people seem to think that DSPAM is either feature-complete and bug-free, or they just lack the motivation/energy/time/whatever to contribute anything.
Because I care about DSPAM, in the years I've active within the project all of these issues have bothered me multiple times. I've tried to address stuff that I could address, or tried to raise awareness for other things. I didn't get much done though.
- From this I can only conclude that DSPAM is a sort of dead project. Just like I wouldn't like to get myself (or my company, for instance) depending on a piece of software in this state, I think that nobody should get the impression that DSPAM is supported by an active community, when that support actually never goes beyond answering easy questions on the users mailing list.
Therefore I'm resigning as DSPAM release manager, and will remove myself as project admin from the sourceforge project. I hope others will take up responsibility for the work I leave behind (and unattended tasks listed above, or any tasks I didn't point out), so maybe DPAM will become an active project in the future.
Kind regards, Tom Hendrikx
PS: I tried to address some of the issues listed above (logging of regular operation, transparent code (including documentation), easier ways of contribution) outside of the DSPAM code base itself, which finally resulted in dspam-milter[1]. I'll continue to develop and support that software, probably until I'll stop using DSPAM myself. If I would ever abandon the project, proper notice (just like this one) will be given.
[1] https://github.com/whyscream/dspam-milter
Open source business process management suite built on Java and Eclipse Turn processes into business applications with Bonita BPM Community Edition Quickly connect people, data, and systems into organized workflows Winner of BOSSIE, CODIE, OW2 and Gartner awards http://p.sf.net/sfu/Bonitasoft
Dspam-user mailing list Dspam-user@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/dspam-user