I'd like to request stabilization of the following packages on amd64: x11-misc/googleearth-5.1.3533.1731-r1 app-emulation/emul-linux-x86-qtlibs-20091231-r1 Google earth has been working well for me for the past few months on amd64, I'd like to stabilize it soon, since it's the latest official version. I've actually been using it with the 2010* versions of emul-linux-x86. However, I've downgraded to the latest stabilized 2009 versions today and run googleearth against them with no problems. Googleearth does depend on a currently unstable version of x86-qtlibs, which was bumped in response to bug #305825(RESOLVED), which looks to be in response to googlearth bug #303789(RESOLVED). Currently outstanding bugs on googleearth: 212373 - Bundled libs. This appears to be an ongoing issue. 251713 - Crash in v4.2 (current stable build) - 5.1 likely fixes this 262780 - Issue with qtlibs localization - 5.1 fixes 265371 - Googleearth preserves libs. File in question isn't even on my machine. The only real blocker I see is the bundled libs, which has a tracker. app-emulation/emul-linux-x86-qtlibs-20091231-r1 has no open bugs at the moment, and the latest r1 fixes a reasonably important flaw in the current stable version (see #305825), so I think it's ok to request that this be stabilized as well. Reproducible: Always
Very nice summary, thanks. (In reply to comment #0) > Googleearth does depend on a currently unstable version of x86-qtlibs, which > was bumped in response to bug #305825(RESOLVED), which looks to be in response > to googlearth bug #303789(RESOLVED). Right. The bump just added an env file which is crucial for the package to work as it is, so I don't think that needs to wait 30 days (also since there's a newer version, no ~arch user tests this anymore). > Currently outstanding bugs on googleearth: > > 212373 - Bundled libs. This appears to be an ongoing issue. Only a few left, certainly much better than current stable. > 251713 - Crash in v4.2 (current stable build) - 5.1 likely fixes this Yeah 4.2 is ancient. > 262780 - Issue with qtlibs localization - 5.1 fixes Right. > 265371 - Googleearth preserves libs. File in question isn't even on my > machine. That's a bug in p.masked portage. > The only real blocker I see is the bundled libs, which has a tracker. I see a different problem - upstream can update the distfile any time (not versioned), so users who will try to install this version afterwards will have broken manifest and I will have to bump it in ~arch again. So far google has provided stable distfile URL only for the last build of each major release, and only after builds for the next release started to appear. The license does not allow mirroring. So the options are: 1) stabilize 5.1 and if it's updated, stabilize new build quickly - no problem for me, minor bumps are usually just bug fixes, but arch teams might not like. On the other hand, there was no new build since november, so the next one could be a beta of next major release (and at that point there should be stable URL for 5.1, but no guarantees). 2) backport the fixes from 5.1 (especially unbundled qt) to 5.0 and stabilize that - some extra work, and probably no ~arch user will be testing it. 3) Just drop 4.2 and have no stable.
I incline to drop the stable keyword, but it is such a popular application...so go with 5.1 and cope with broken Manifest from time to time.
(In reply to comment #2) > I incline to drop the stable keyword, but it is such a popular application...so > go with 5.1 and cope with broken Manifest from time to time. Right, so please do :)
amd64/x86 stable, all arches done.
