ALERT: KDElibs (at least) 4.8.4 is actually 4.8.80+

[Albert Astals Cid]

El Diumenge, 10 de juny de 2012, a les 22:24:26, Andreas K. Huettel va 
escriure:
> Am Sonntag 10 Juni 2012, 15:27:55 schrieb Albert Astals Cid:
> > > - - Before announcing the tarballs, build the whole thing at least once.
> > 
> > We do that, that's why we have a week before the release where packagers
> > get access to pre-release tarballs.
> 
> I hope you're not too serious about that. None of us distribution packagers
> is actually using the unmodified build scripts. So there are lots of
> additional problem sources. Yes they are our problems then, but it would be
> helpful to know that "it could work". (It is not hard. Take one reasonably
> fast box with some sane average defaults. Install what (according to
> release notes) should be installed. make install. Come back in a few hours
> and inform whoever's module failed.)
> 
> That aside, I think a few build failures in a >pre-release< >beta< tarball
> the size of KDE should not be a problem. After all, we want to help, and we
> still have some time figuring out solutions to problems.
> 
> What I am worrying much more about is
> 
> 1) confusion about dependencies (Soprano, SDO, Qt): e.g, in an ideal world,
> there would be a dependency freeze some time before the release, and it
> would only be allowed to use dependencies actually released publically at
> *that* time.

http://techbase.kde.org/Schedules/KDE4/4.9_Release_Schedule#Thursday.2C_May_10.2C_2012:_KDE_4.9_Dependency_Freeze

The problem is we have no ways of enforcing it, can't fire people, can't send 
them professional assassins, so if people don't follow it, all we can do is 
send mails, which is what we are doing.

Cheers,
  Albert

> 2) my feeling of a different perception of "stable" between packagers and
> developers