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

Re: A new arch support proposal, hopefully consensual (?)



On Sun, Mar 20, 2005 at 08:48:45PM +0100, Julien BLACHE wrote:
> Raphael Hertzog <raphael@ouaza.com> wrote:
> 
> > I believe everyone is supportive of the various ports, nobody has any
> > interest in making a port fail... but it's clear that many maintainers
> > are frustrated to be blocked because their package doesn't build on an
> > arch they don't care about.
> 
> They should care about every architecture we support. Because our
> architecture coverage is an important part of the spirit of this
> Project, and an important feature.

I've been thinking about your post for a couple of days. I think that
for some people, multi-platform support is important but everyone has
different reasons for being involved and different opinions on what is
important.

Our users and developers might want a distribution which

1. Runs on every platform; OR
2. Is 100% free, but only needs to run on their mainstream desktop; OR
3. Is technically the best at any cost; OR
4. Suitable for production use on modern hardware; OR
some combination of these etc.

It's unfair to expect everyone to share your goals and in particular
your desire to port to any particular platform. None of these goals is
more correct than the others, though you might say that currently
#1 is dominating #4 because it's slowing down the release of sarge.

Debian wasn't originally multi-platform. The second platform was m68k,
added in 2.0. The original Debian manifesto talks about building a
technically excellent product and building a community to develop a
distribution together. So please don't expect that everyone shares 
your interests.
( http://www.debian.org/doc/manuals/project-history/ap-manifesto.en.html )

In my reading, the proposal says that the release, ftp and post-release
teams don't have the resources to concentrate on every platform, so if
you want to see it maintained and released, you have to help.

> As has been noted already, having a 18-24 months release cycle isn't
> much of a problem; actually, enterprise users pretty much like that,
> although 12 months is a more generally acceptable timeframe.

Right, but we're at 32 months now and counting.

Hamish
-- 
Hamish Moffatt VK3SB <hamish@debian.org> <hamish@cloud.net.au>



Reply to: