What makes a spin a Spin?
mmcgrath at redhat.com
Tue Feb 26 20:17:34 UTC 2008
On Tue, 26 Feb 2008, Josh Boyer wrote:
> Um, updates for what? Anyway, you're the IS guy so I'm not going to
> argue on the overhead, particularly when you've been so accommodating
> thus far.
They're F8 boxes:
Install 1 Package(s)
Update 38 Package(s)
Remove 1 Package(s)
Recently installed already have updates, that need to be tested, etc, blah
> > what? Because we're making another excuse for SELinux? /me not a fan of
> > that. There's a right way to make spins, its on the builders in a chroot.
> I won't disagree, though Jeremy certainly has more technical details as
> to why that's hard.
> I will note, for those that aren't aware, that we requested the guests
> to leverage the faster connection speeds that are available there, as
> well as having a controlled environment. Speaking from that angle,
> they have been invaluable. So having _some_ kind of machine, builder
> or not, is awesome.
I'm just worried that because a quirk in selinux is hard to fix, the
burden of this problem got shifted onto the Infrastructure Team and will
be forgotten about because "spins are getting made".
If the selinux problem isn't going to get fixed at all the infrastructure
team needs to know that so we can make a permanent solution, include it
in our budget, create the SOP's for the machines, monitor them when
required, etc, etc. The temporary fix we did make was so the spins
could continue going through.
More information about the fedora-advisory-board