Re: petsc_2.3.0-1_i386.changes REJECTED
On Mon, 2005-11-14 at 20:56 +0100, Thomas Viehmann wrote:
> Adam C Powell IV wrote:
> > As you could see in my message, I did ask Joerg again, but with no reply
> > for a week, so nobody but Joerg knows what I did or didn't convince him
> > of.
> Well, I'm not Joerg, but based on the evidence of rejects/accepts and my
> understanding of policy and actual practices, it is my understanding that
> - the meta packages probably are the main rejection cause,
> - parallel installability of -dev-Packages is not an issue Debian cares
> about except for the most popular packages,
> - (this doesn't directly relate to the rejection, but nonetheless) the
> use of the alternatives system is probably a bad idea because it
> leads to unexpected compilation results (essentially many developers
> considere it a bug for a package build to produce substantially
> different builds when run on arbitrarily chosen current unstable
> - versioned -dev package naming is not Joerg's preference, but when I
> asked him about a rejection of libfoo12-dev and pointed out Junichi's
> libpkg-guide recommendation he said he wouldn't reject for this in
> the future.
This is good to know, thanks very much for the clarifications. It's
unfortunate that Debian doesn't care about parallel installability of
-dev packages, given how many user communities could likely benefit
(PETSc's was started at the request of several users).
> P.S.: Should we move elsewhere from -release?
Sure, Joerg and I are discussing this. Seemed like the appropriate
place to ask when I hadn't heard any clarification in over a week.
GPG fingerprint: D54D 1AEE B11C CE9B A02B C5DD 526F 01E8 564E E4B6
Welcome to the best software in the world today cafe!