Refining the update queues/process [Was: Worthless updates]

[James Antill]

On Wed, 2010-03-03 at 18:06 +0100, Ralf Corsepius wrote:
> On 03/03/2010 05:10 PM, Seth Vidal wrote:
> >
> >
> > On Wed, 3 Mar 2010, Peter Lemenkov wrote:
> >> And what about tickets, closed with "FIXED UPSTREAM" w/o actually
> >> applying fix to a package?
> >>
> >
> > Those items will be released in the next release of yum or in the next
> > fedora release. If the bug is serious or a security issue I will often
> > backport a patch.
> 
> QED: You don't fix your bugs, but prefer confronting users with them and 
> force them to dig out your "fixes" to work-around the issues they are 
> facing.

 No, we don't fix _all_ bugs immediately, so as to not introduce
unnecessary regressions. This is the engineering part of "software
engineering". And, yes, users can often personally work around their
problems by installing from rawhide/etc. ... without adding risk to
anyone else.
 Some bug reporters have argued that their bug is more critical than we
think, some do change our minds, that's fine and I'm happy to discuss it
with them. But your desire that we should instantly rebase for all bugs,
immediately, and no matter the risk is something you will have to live
without.

 Again, rawhide is => that way. Have fun.

-- 
James Antill - james at fedoraproject.org
http://yum.baseurl.org/wiki/releases
http://yum.baseurl.org/wiki/whatsnew/3.2.27
http://yum.baseurl.org/wiki/YumMultipleMachineCaching