Re: Weirdness in "apt-get upgrade"
On Fri, May 31, 2002 at 12:06:48PM -0700, Brian Nelson wrote:
> Colin Watson <firstname.lastname@example.org> writes:
> > Yes. I've had to do the odd security update of my own packages in the
> > past, and I had to build packages for every architecture by hand.
> > Finding Debian-administered machines of the right architectures on which
> > the right build-dependencies are installed is a pain - and that was on
> > six architectures rather than the eleven that are going to release with
> > woody. (In fact, back then I couldn't get access to a suitable m68k
> > system at all, and I had to wait for one of the security team to sort
> > that out for me.)
> I'm curious if there are any usage statistics out there regarding
> Debian's ports. I would guess that at least 99% of all Debian users
> use maybe 2 of those ports. Who here has even heard of SuperH?
We use five at work, if I've counted right, and I have several friends
who between them use two or three more. (Maybe my friends are atypical.)
SuperH isn't really relevant as it isn't being released.
99%, perhaps ... but for the remaining 1% it's invaluable, and many of
those 1% are valued developers.
> It seems silly that Debian spends so much time and effort on so many
> ports, and consequently delaying the stable release for months, when
> so few people actually use the ports. If it takes so long and is such
> a pain to build packages on ancient archs like m68k, why not just
> release i386 now and release the other ports later when they're ready?
Debian is the only distribution that supports a number of these
architectures at all, and we're their last port of call; the commercial
distributions won't touch them if they don't improve the bottom line. As
a result there's a very strong movement within Debian not to treat
non-i386 users as second-class citizens.
The problems with building packages for lots of architectures in
unstable have largely been solved, anyway, and are being solved for
stable. From potato to woody there have been massive improvements in
both infrastructure (new archive maintenance software, 'testing',
distributed build daemons) and packaging (build-dependencies) which are
making it easier to support multiple architectures for the future.
It's also worth pointing out that the effort that some people see as
being wasted on other ports actually benefits the distribution as a
whole in the long run. For example, somebody complained a while back
about the number of bugs filed because packages didn't build on hppa;
what I suspect he didn't realize is that a large chunk of those bugs
were about gcc 3 support in C++ packages, which will be i386's default
compiler soon! It starts seeming a lot more worthwhile even to i386-only
people at that point.
Fundamentally, Debian is committed to being a multi-architecture system
for as long as porters are willing to support it. The Project Leader
posted to debian-devel-announce about porting a couple of weeks ago, and
summed up the issues quite nicely.
Colin Watson [email@example.com]
To UNSUBSCRIBE, email to firstname.lastname@example.org
with a subject of "unsubscribe". Trouble? Contact email@example.com