Heads up! Broken deps in Upgrade from 12 to 13

Braden McDaniel braden at endoframe.com
Sun Feb 21 04:34:39 UTC 2010


On Sat, 2010-02-20 at 20:20 -0800, Conrad Meyer wrote: 
> On Saturday 20 February 2010 06:03:28 pm Braden McDaniel wrote:
> > I guess I should properly read the headings... I guess I misunderstood
> > where this was going from your summary.
> > 
> > If I understand it correctly, it's telling me that a F12 that hasn't
> > been updated will have problems being upgraded to F13.  Certainly that's
> > not ideal; but it doesn't strike me as exactly tragic, either.  How
> > important is this considered, generally?
> > 
> > I'm not crystal clear on why this is broken, either.  It may have
> > something to do with the fact that the openvrml binary RPM changed to
> > being a metapackage; but the implication of Boost is a bit confounding.
> 
> The ratio of "badness" to the difficulty of fixing it is fairly high. Fixing 
> it is fairly trivial. Leaving it broken breaks upgrades for users who upgrade 
> distributions using yum update -- which, though it isn't supported, is 
> reasonably common.

I'm happy to fix it if I can figure out what's broken.

> The solution varies somewhat depending on how broken it is; I'll try to 
> describe all possible problems / solutions. There's going to be a lot of text 
> here, but nothing hard to do.
> 
> If the F-12 package has a higher Epoch than the F-13 package, this is *very* 
> broken, but can be fixed by bumping the F-13 Epoch to at least the F-12 Epoch.

That's not it.

> If  the F-12 package has a higher Version than the F-13 package and the same 
> Epoch, this is probably broken.

Also not it.

> If the F-12 package has the same-or-lower Epoch and the same version, but a 
> higher Release, then a couple things may be wrong:

Nope.

> I think that's everything. Let me know if you have any questions, or if I'm 
> wrong ;).

I'm afraid I'm still at a loss.

-- 
Braden McDaniel <braden at endoframe.com>



More information about the devel mailing list