brad at stop.mail-abuse.org
Thu Mar 17 02:05:08 UTC 2005
At 11:03 PM +0000 2005-03-16, Per Hedeland wrote:
> It's quite clear that the
> mailman developers think that what they're doing is the Right Thing, and
> even though I obviously disagree, it's their software and they decide
> how it should work.
I count myself as one of the Mailman developers. I help maintain
the mailman-users and mailman-developers mailing lists, and the
overall mail system for python.org. I've got full root access to all
the affected machines. I've made a few minor hacks to Mailman and
submitted them via the patch tracker. I'm also on various internal
mailing lists related to the python.org site, Mailman in general,
I'm not Barry, or Chuq, or Tokio. I'm not one of the core
developers. I'm not a part of the internal cabal that gets woken up
at 2AM when there's a new zero-day security hole that has been found.
But I do consider myself to be a member of the greater Mailman
Speaking as one of the Mailman developers, I feel that I can
safely say that we are open to any improvements you may have. In
this particular case, the code that we have today is what we have
arrived at through hard lessons learned through unpleasant direct
experience. I've got quite a bit of experience as a news
administrator at a top-rated European site, but I'm not a programmer,
and it's hard for me to judge the overall technical capabilities and
deficiencies of code in this area.
However, if you have something that will solve the problem
better, then more power to you. In that case, please feel free to
upload your suggested code changes to the patch tracker, and let me
know once you have. Tokio (the Release Engineer for Mailman 2.x) has
already done a couple of things on special request for me (both of
which are being incorporated into the upcoming 2.1.6 release), and I
imagine that I'll be able to get him to take a look at your
submission in a reasonably short period of time.
> All I want is to get a usable comp.protocols.time.ntp back, and I'm
> prepared to help with achieving that limited goal. Since I gathered that
> using mailman's builtin functionality for the gatewaying was a given,
> the obvious solution was to do the needed modifications to the copy of
> the mailman software that is being used for this gatewaying.
You're not going to get any changes made to the mailing list
management system we run, without going through the proper Mailman QA
process. I'll muck about with whatever code on a test system, but
I'm not going to do risky things on a production system. Part of
that process means that the code has to be approved by Tokio, or he
has to create his own code based on the concept.
If the code change is significant, then I'm not going to run with
a patched system -- you're going to have to wait until the code in
question gets included in an official release version.
> But of course local mods can be a bit of a headache maintenance-wise
> (though not much if you keep the softare in e.g. CVS, which has specific
> support for such things) - so maybe another alternative could be to just
> let mailman manage the mailing list, and do the gatewaying outside of
That offer has been made before, but I trust other mail-to-news
gateways less than I do Mailman. We have known limitations and
issues, but they're known. I don't trust anyone else's code in this
area as far as I can bodily throw their planet of residence.
Brad Knowles, <brad at stop.mail-abuse.org>
"Those who would give up essential Liberty, to purchase a little
temporary Safety, deserve neither Liberty nor Safety."
-- Benjamin Franklin (1706-1790), reply of the Pennsylvania
Assembly to the Governor, November 11, 1755
SAGE member since 1995. See <http://www.sage.org/> for more info.
More information about the questions