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

Re: [Debconf-discuss] Talk selection for Debconf 7

On 9 Jun 2006, Christian Perrier told this:

> Quoting Manoj Srivastava (srivasta@acm.org):
>> As I said, timing could be worked on. At least the people
>> presenting the talks know they are coming, and a number of people
>> not giving talks are more or less committed by that time. However,
>> as long as the people believe there is a possibility they shall
>> attend, then the market principle shall hold.  Indeed, if we
>> encourage people to register if they think there is a possibility
>> they could come, and confirm later,
> Yeah, that could be possible. I however fear that this would end up
> in a "love/hate" game as we are obviously very good at it in Debian
> (not only Debian, but we seem to emphasize that kind of behaviour).
> As you mentioned, the risk is seeing "popular speakers" being
> overrated whatever they might propose.

        I have two observations on this point. Firstly, if we were
 selecting _one_ talk, I would agree that personal bias could affect
 the outcome. But we are talkging about rating dozens of proposals,
 and selecting dozens still.  I think you might be surprised at how
 well market forces overcome individual biases when taken in bulk.

        Secondly, I am not sure that going against the martket
 decision is wise, even then: popular debconf speakers do not gain
 popularity because they look like movie stars; they are popular
 because they have done seminal work in Debian (and have thus things
 to convey to the rest of us), have demonstrated sound technical
 judgement, or have demonstrated in the past that their talks are fun
 and instructuve.

        Why should we try to put a handicap on these talks? Why isit
 not possible that these people are popular because their talks are
 likely to be good? What justification do you have to show that
 "popular" speakers, if selected for, actually harm the confenrence?
 Surely we have enough slots that good speakers with a track record of
 informative talks do not crowd out deserving candidates?

> I would say currently that a voting system could be one of the
> rating factors, while there is still an academic comitee in charge
> of doing the final selection, with the ranking as one of its
> input...other input being their wisdom...
        This smacks of the soviet era central planning, really. The
 politburo thought that a few experts would outperform random
 investments made by the unwashed, and often decadent uneducated
 western capitalists, since they had a plan, had education, tools, and
 sought "balance".  The market outperformed them at every turn.

> This academic comitee certainly needs the best recognition it might
> have just to make its decisions widely accepted. Actually, for DC6,
> they have been quite generally properly accepted and, as far as I'm
> aware of, there haven't been many complaints about their choices.

        My experiences differ (I have talked with several people with
 profound misgivings) -- but this is about making the future better,
 not about recriminations.

It may soon be time for you to look for a new line of work.
Manoj Srivastava   <srivasta@acm.org>  <http://www.datasync.com/%7Esrivasta/>
1024D/BF24424C print 4966 F272 D093 B493 410B  924B 21BA DABB BF24 424C

Reply to: