Tristan Santore wrote:
All very nice and all, that you are porting this, however, why not
contribute to F17 and beyond, because then we might/will end up with a
proper RHEL/CentOS ARM version sooner.
Maybe, maybe not. I am not as convinced as you. There doesn't appear to
be any sign of a race to support extra platforms. The RHEL clones
(CentOS, Scientific) only support the primary architecture (x86).
If two people can do this, working on it as and when time permits over a
3 month period in the evenings and on weekends, then it seems most
implausible that a company with a billion dollars a year in revenues
employing hundreds of competent engineers didn't do this because they
couldn't. The only plausible explanation I can come up with is that they
didn't bother because they saw no gain in doing it.
If you want to add ARM support to CentOS 6 or Scientific 6 - go for it.
I did all the hard work - all you have to do is find all the packages
that I changed (they have a postfix .0 in the version after the distro
tag, so very easy to find) and rebuild CentOS complete with it's
branding and you're good to go. However, since CentOS and Scientific
only support x86 even though technically RHEL also supports PPC, I
didn't think it particularly likely that they would be interested in
supporting another platform, especially since that involves additional
maintenance work (which PPC in theory wouldn't since it is maintained
upstream).
There has been a buzz in the past
about low power usage, performance multi-core arm chips, which can be
stacked in racks.
Indeed - I did this because I wanted a distribution that would be
maintained longer than Fedora for my own ARM machines - even if I have
to do package maintenance myself, having an upstream that is maintained
with care to not break things too badly (at least most of the time) is
useful. If this is useful to you, too, then by all means, take my
modified packages and do with them as you please. :)
Gordan