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

What we can learn from Canonical/Ubuntu

#include <hallo.h>
* Josselin Mouette [Sun, Jun 05 2005, 06:32:10PM]:
> Le dimanche 05 juin 2005 à 18:20 +0200, Wouter Verhelst a écrit :

> This is a real question, that deserves something else than "there is no
> cabal" as an answer. Debian has succeeded so far as the project was
> independent from any institution or company. If this changes, we're
> going to a dead end.

Really? IMO it is exactly the lack of authority and strong top
management that has lead us into the current situation.

Take it as is: Woody release was a real misery, and Sarge's close to a
fiaco (sligthly exxagerating). What's the reason? IMO simple: we cannot
synchronise our actions properly. This happens because of missing
motivation, but this is a side effect of the real reason: having no
concrete tasks, no concreate plans or scedules. And no authorities that
would give the answers (even worse, official or pseudo authorities
hiding and avoiding to have to answer, you know whom I am talking about).

The dead end expects Debian because of people not wanting to take
responsibility and people that act irrational (eg. too risky in task
scheduling issues, time calculations) because they know they can afford
it, not having to take any responsibility.

Are you jealous of Ubuntu because they have more success then Debian?
Why about seeing them as a software experiment and learning from them?
We would learn following things:

 - a hard release schedule is neccessary. Everyone needs to know _when_
   the job is to done, there must be a clear deadline. Even if there is
   just small leeway for delay, people will be unable to schedule their
   work. We have seen that before, again and again, even AJ has
   complained about that. We should stop compensating this shit with
   stretching the release period, again and again.
 - we need to priorize packages and architectures. "Second class" may
   sound bad we need it to make the core system thin enough to become
   release in timely manner. Keep the best, drop the rest (or move it to
   separated archives).
We cannot have every piece of software in the world in Debian and
guarantee a high quality level. Remember the numbers of upstream bugs
that we have in our "stable" all the time.  After view months it simply
outnumbers the potential risks from having an upgrade.
So recommending Debian Stable as production system has become pure farce
in the last months, because it is either buggy (upstream bugs) or
incompatible == unuseable because of hardware issues (outdated software)
or compatibility with newer developments.

What I mean: Debian has become a no-no for productive environment for
the majority of users out there (and please do not come with "works for
me on ...", it is not very representative).

Ein reicher Mann ist oft nur ein armer Mann mit sehr viel Geld.

Attachment: signature.asc
Description: Digital signature

Reply to: