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

Re: DPLs : what do you think about ...

On Mon, Dec 14, 1998 at 11:35:17AM -0800, Joey Hess wrote:
> Ben Collins wrote:
> > You fail to see that the freeze would be shortened if it were handled
> > correctly. Which means your updates would not be held off long.
> No, I don't fail to see it. I disagree that this change would shorten the
> freeze appreciably, because I disagree with the premise of your whole
> argument, which seems to be that we developers are short-sighted people who
> can't concentrate on getting one release out the door without being
> distracted by unstable.
> Even if you're right and we could reduce freezes to a month, a month is an
> unacceptable stall for me on developing my debian-only packages.

I honestly fail to see how you can't develop packages without unstable. If
the reason is that unstable is your release method for non-debian specific
programs, then I disagree with it completely. If it is because the
packages are debian utilities, then I fail to see how you cannot still
work on them and wait till after the freeze in order to introduce new
features (since bug fixes would obviously go into the freeze).

Also, the fact that good proposals fall into the cracks release after
release and things that need to be done seem to fall by the way side,
shows that not enough attention is being given to detail. This is not a
blanket statement about all of the maintainers, but is the about the
development cycle in general. How many times to you read on -devel "This
has been proposed before and there was little against it, but nothing ever
became of it".

There are alot of ppl who would like to see "proposal A" be implemented,
but since we are constantly being driven by the "package" mindset, these
things usually get little support and are never realized.

It is my contention that taking a break from the full development cycle
would allow for things to settle down long enough to regroup some of our
goals and efforts in order to complete some tasks that would usually never
get done. 

I truely hope that no one thinks this is an attack on any specific
developer or the way things are done now. If so, I apologize. My only
reason for such as proposal was that it seemed to be a clear way to gain
some order in the constant flow of updates to an unstable fork. There are
actually some maintainers who would welcome a small break in this full
cycle. And I'm sure that there are some who are unable to keep up with
this constant flow because of connection problems, events in their
personal lives, etc...which they would be able to keep up with given this

Remember we are not bound by any release dates, why should we rush to have
another full development cycle when such an easy change could improve what
we are doing? 

I can't drop this idea since there are so many benefits from it and I have
seen very few real problems with it.


-----    -- - -------- --------- ----  -------  -----  - - ---   --------
Ben Collins <b.m.collins@larc.nasa.gov>                  Debian GNU/Linux
UnixGroup Admin - Jordan Systems Inc.                 bcollins@debian.org
------ -- ----- - - -------   ------- -- The Choice of the GNU Generation

Reply to: