[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: [Debconf-discuss] Registration FAQ

Gunnar Wolf <gwolf@gwolf.org> wrote:
> MJ Ray dijo [Tue, Feb 26, 2008 at 02:32:47PM +0000]:
> > The lack of topic grouping is a bug in debconf set-up, isn't it?
> There are too many vectors around which you could end up grouping
> talks - it's not as trivial as it looks like. [...]

I agree with that in general.  I spent time studying operations
research during at university level and I've helped to organise
various conferences.  But "too many"?  Is better talk scheduling
really impossible here?  "Painfully many" perhaps?

> Of course, I do not know if such a subsystem would end up being useful
> in the general overview - Would it really be better than
> essentialy-random-allocation?

Yes, I believe it should be, in several ways.  If done in at least two
stages with up-to-date talk submission statistics, it could even
encourage better talks on popular specialist-knowledge topics and
discourage talks on some unpopular widely-known topics, because people
wouldn't even bother submitting a weak proposal for a competitive
subject, or talks in languages where no talks will be accepted.

The scheduler subsystem could also prepare people better for any
talk's rejection, (by showing proposals to time available ratios), or
a session getting dropped or shortened (if its supporters/time drops
less than twice that of another session) or do any number of things.
I've set out my priority (which days should I attend?) and argued for
its relevance and importance but what are the other desired aims?

> How compatible would it be (without
> major and inconvenient adjustments) with human-induced changes
> (i.e. "I'm scheduled for Monday, but I arrive on Tuesday", or "We all
> know $person always fills up any room he gets - Schedule him with no
> competing talks!")... Human factors induce stupid amounts of
> complexity and frustration.

It depends how it's done, but I think hard constraints like that are
simpler to handle than preferences in most systems.

> [...] Debconf is meant for work in and towards
> Debian. We do not need a 500+ people conference! IMHO, we would lose
> more than what we'd win.

Is 500+ too much?  IIRC, debian has around 1000 developers and at
least 3000 contributors.  Debconf.org claims "DebConf is the Debian
Project's developer conference. In addition to a full schedule of
technical, social and policy talks, DebConf provides an opportunity
for developers, contributors and other interested people to meet in
person and work together more closely." If that's no longer the aim,
it's time to update the descriptions and invite the fork.

Hope that helps,
MJ Ray http://mjr.towers.org.uk/email.html tel:+44-844-4437-237 -
Webmaster-developer, statistician, sysadmin, online shop builder,
consumer and workers co-operative member http://www.ttllp.co.uk/ -
Writing on koha, debian, sat TV, Kewstoke http://mjr.towers.org.uk/

Reply to: