Re: etch or later?
On Sun, 2006-06-18 at 19:05, Rene Engelhard wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> Thomas Walter wrote:
> > So, what prevents recompiling this package with a newer version of gcj?
> Nothing. That wasn't the problem. That bug is quite new. The problem is
> *neon* blocking transition into testing currently (and a missing
> powerpc build). Now, when subversion/perl are ready it will go in, but
> OOo not - because it currently has conflicting build-deps (mozilla vs.
> xulrunner which *will* be fixed, but not anymore for 2.0.2, but for the
> next upload which will be 2.0.3rc5 or rc6) but because of that I can't
> provide a powerpc package for 2.0.2-3 which is a prerequisite for
> 2.0.2-3 entering testing.
> That RC bug (which is not really RC anyway after all can be ignored for
> etch atm, that'd not a big problem)
Thank you for these details.
This and all the other items related to OO show that this large
application is not a good example.
Due to its size, the different kinds of components and the large
requirements OO makes to libraries, compiler, interpreters (java), ...
makes OO to an extra ordinary package.
Maybe one can define exceptions for these kinds of large packages?
>From your description above, I understand that for PPC, OO 2.0.2-3 has
no chance, but the next release.
So, making an exception and kick 2.0.2-3 into testing would be a natural
choice from my point of view.