[Fwd: Question: Will Fedora 10 Contain KDE 3.5.10 or Not?]

seth vidal skvidal at fedoraproject.org
Wed Aug 6 16:51:08 UTC 2008


On Wed, 2008-08-06 at 16:42 +0000, Paul W. Frields wrote:
> On Wed, 2008-08-06 at 18:36 +0200, Thorsten Leemhuis wrote:
> > On 06.08.2008 15:05, Jesse Keating wrote:
> > > On Wed, 2008-08-06 at 12:27 +0200, Michael Schwendt wrote:
> >  >
> > > The "too many updates" problem is something I've been trying to word so
> > > that others share my opinion that something is wrong here.  I haven't
> > > been able to effectively communicate what I perceive to be a problem. [...]
> > 
> > My 2 cent: the number of updates is not the problem. In fact lots of new 
> > an up2date software without being to bleeding edge (like rawhide) is 
> > exactly one of the things that make Fedora great imho.
> [...snip...]
> 
> > Level 1 -- rawhide, similar to how it is today (a bit more stable and 
> > less breakage would be nice, but that's in the works already)
> > 
> > Level 2pre -- things that got tested in rawhide, that are still young, 
> > but known to work well in rawhide; similar to what updates-testing for 
> > F9 is today;
> > 
> > Level 2 -- things that worked fine for some time in 2pre; similar to 
> > what F9 is today
> > 
> > Level 3pre -- things that worked fine for some time in 2
> > 
> > Level 3 -- things that worked fine for some time in 2pre
> > 
> > 
> > Level 3pre and 3 are like F8-updates-testing and F8, but with the 
> > difference that everything has to be tested and shipped in level 2 (aka 
> > F9) first.
> 
> Interestingly, this is is sort of what Seth Vidal recently did for yum
> -- kinks were worked out in upstream and Rawhide, he has done several
> useful updates for F-9, and only recently has he bundled it up for an
> F-8 update.  Hopefully he'll chime in with his thoughts on this
> strategy.
> 

I think doing the above for every package creates the problem that jesse
is talking about.

It seems like what jesse has talked about in the past is trying to get
package maintainers to realize that we don't need to issue an update for
utterly minor fixes. You can collect a set of things and push them back
in a batch.

The problem is this is based on Jesse's feeling that we're issuing a
bunch of trivial updates and we're having trouble getting a metric on
how accurate his feeling is.

Also the reason why we're conservative on yum updates w/older releases
is simple: if yum breaks the ability to get more updates breaks too,
which is a big problem.

-sv

-- 
I only speak for me.




More information about the fedora-advisory-board mailing list