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

Re: Debian needs more buildds. It has offers. They aren't being accepted.



On Thu, Feb 12, 2004 at 02:17:54PM +0000, Martin Michlmayr - Debian Project Leader wrote:
> * Chris Cheney <ccheney@cheney.cx> [2004-02-12 01:06]:
> > Saying "oh well" about lack of communication is not something I
> > would expect to hear from DPL!
> 
> Oh well.  Debian is a very diverse community and there is no way
> everyone is going to be able to communicate with everyone.  This is
> just a fact I have to acknowledge.  There are various possible
> solutions from this.  One would be to make people more communicative,
> but this is usually not possible because people have their personality
> and won't change.  The other possibility would be to replace the
> people who cannot communicate and replace them with people who can
> communicate _and_ do the work.  Unfortunately, finding such people is
> often hard (there are some good examples in Debian for this, though,
> like Colin Watson just to pick one who is a sheer pleasure to talk to
> and who gets lots of work done).  In the specific case of buildd, Ryan
> has done the major work to keep the buildd infrastructure running in
> the last months, and there is no better replacement.  He could be
> replaced by someone who is more communicative, but then the work
> wouldn't get done.  Is that what you want?  The other option, and the
> one most appropriate in this case imho, is for the DPL to act as
> intermediate.  I don't say "oh well, there is a lack of communication,
> let's just not talk about it and ignore this issue.  Instead, I say
> "oh well, I cannot change the personality of these people".  However,
> something I _can_ do, and this is one reason I decided to run for DPL
> last year, is that I _can_ talk to all kinds of different people.  I
> have no problem talking to Goswin and Ingo, and at the same time I'm
> in virtually daily contact with James and also usually can communicate
> with Ryan.  And, in fact, I found two mails from him in my mbox this
> morning: one about mipsel, one about mips.

I don't doubt that James and Ryan do a lot of behind the scenes work.
However, if they are too busy to communicate or just don't like to then
they should have an assistant that will communicate as Goswin mentioned
in his reply. Communication will make you a little bit less efficient at
_YOUR TASK_ but will allow many others to be more efficient by being
able to organize better. For example with the arm/mipsel problems I
would have known why the buildds weren't building KDE packages for a
month, and could have probably already uploaded KDE 3.2.0 by now.

I do appreciate your update but don't you have many other things that
deserve more attention? A DPL shouldn't have to root around in problems
of subsections of Debian especially for something as petty as people
refusing to communicate. It takes time away from you which could be
better spent representing Debian to others.

> If there is lack of communication, the DPL surely has to deal with it.
> I approach it by stepping in as intermediate and to resolve the problem
> this way, and I think this fulfils the role of the DPL quite well.
> (Other DPLs might approach it differently - it's all a matter of
> personality; in the end, the question is which approach works best,
> and I think mine is quite successful, certainly more successful than
> just replacing various people with others when there are no good
> replacements available.  You can only work with the resources you
> have, and I try to do that as well as possible.)

Perhaps not replace but definitely have assistants, and if the people
also refuse assistants then perhaps its in Debian's best interests to
replace them with someone less experienced who can either communicate or
can work with an assistant who can.

Lack of communication has been a problem in Debian for a very long time.
I know that I would vote for a DPL who would make that one of the top
things they would work to get fixed.

Chris

Attachment: signature.asc
Description: Digital signature


Reply to: