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

Re: Proposal to remove archs

I'm just going to assume that this was posted to -boot to get the d-i
side of the story and respond like that..

Ben Collins wrote:
> I'm not subscribed to debian-boot, so I'll preface this with that fact
> that I have not read any of the discussion so far, so excuse a rehash.
> Simply take this in. I'm not posting to drag out any further debate, only
> to add my two cents.
> I think removing sparc would be a huge mistake. When people like David
> Miller, who currently works for Redhat, have switched to Debian because we
> support this architecture, it means a lot. Davd develops sparc64 using
> Debian now.

I'm really psyched to see him on debian-sparc. That's cool.

> Removing our support for the architecture only opens the door for someone
> else to do so.

I don't anticipate that d-i will remove support for any architecture as
long as we continue to have enough people in d-i to support it. 

For sparc we have Jurij and Joshk doing kernel stuff, Joshk trying to do
silo stuff (but he could use some help), Joshk (sorta) and I (sorta)
doing d-i build stuff, Frans Pop doing some installation report
processing and user support, and various people doing testing. This is
enough people to keep d-i working on sparc, but only just, and when
someone gets busy with other stuff, the sparc port has suffered.

For example right now we've not had daily builds for sparc since Feb
25th, because joshk lacks a machine to run them on and has been busy
with other stuff too I think.

Right now I'm more concerned about the arm port, where we had to pull in
Kenshi to do builds, and while a few people in the d-i team have picked
up the occasional peice (thanks, tbm) there seems to be nobody covering
most of the stuff that needs doing from day to day or whom I can go to
if there's an emergency. hppa has also been problimatical. Jeff has been
doing a great job with the builds and the kernel is in good hands, but
for d-i development and maintainance we've had only Bdale, who has many
other commitments.

What we have had to do a few times in the past, and could conceivably
have to do again in the future is make a d-i release that did not
include a particular architecture, for that particular release, because
there was nobody to work on it. I think the last time was beta3, which
did not include powerpc.

My goals for d-i releases post-sarge have always included continued
frequent d-i releases from testing so that users can easily install that
(or unstable of course). With the SCC plan we will need to look at what
we need to do to make d-i releases for arches in SCC. These releases
might install unstable, or we might work with the arch's porters to do d-i
releases to go with more stable releases of an arch.

see shy jo

Attachment: signature.asc
Description: Digital signature

Reply to: