On Fri, 2007-11-09 at 14:06 -0500, Jesse Keating wrote:
Now that Fedora 8 is out, I'd like to make a suggestion with
PPC. It's no secret that we've been talking about dropping PPC. First
it was all together, then it was so that it could be come a secondary
arch. So far, we've failed to execute.
Digressing from the technical side, I'd like to properly understand your
motivation. _Why_ do you want to do this?
I like Fedora. I like to see it expand and get used in places it
couldn't have been used before. I've always wanted to see it ported to
new architectures and made versatile enough to do new things -- to be
used as a basis for more than the bog-standard Linux desktop/small
server on a PeeCee.
I'd love to see more functionality -- more _possibilities_ -- merged
under the 'Fedora' umbrella.
As far as I can tell, you have the opposite desire. You seem to want to
reduce what we have -- even when something's _working_, you want to back
away from it -- disown it and all but abandon it. And even 'advertise'
that we're doing so -- you want to shout it from the rooftops that we're
backing into our corner.
We strive to ensure that PowerPC "Just Works™", when it comes to QA and
release time. I think we do a reasonably good job, don't we? Stuff slips
through occasionally, of course -- we should do more testing with
SELinux enabled, for example. But in general, I don't think we do badly.
So why do you constantly fight to throw out our work? What benefit do
you think this attitude brings to Fedora? I assume you don't _just_ do
it because you enjoy the massive slap in the face that we feel every
time you do it? :)
Have I missed a large amount of work which you personally need to do for
each release? Or missed a large store of 'private' bugs in bugzilla
which have never come to my attention? Or just misinterpreted the state
of Fedora on PowerPC despite the fact that I use it on all my primary
So what _is_ the motivation?
If Fedora/PowerPC is not good enough to be called 'Fedora', I will
heartily agree that we should stop calling it that. That isn't the case
If Fedora/PowerPC causes a significant amount of work for you, we can
strive to take that load off you -- although I understood that's part of
what the "secondary architecture infrastructure" was _for_, and we'd
agreed that reclassifying Fedora/PowerPC would wait until that was in
It can't be _just_ about the number of users -- I bet that there are
fewer users of Kerberos on Fedora than there are PowerPC users, but we
don't get people ranting after every release that they want to drop
Kerberos support. Our usage numbers are very tenuous anyway, and we
certainly can't factor in the bizarre things like the fact that Linus
chose Fedora _because_ it runs on PowerPC.