[Debian-med-packaging] r-other-mott-happy_2.3-1_amd64.changes REJECTED

[Charles Plessy]

Le Sat, Jul 20, 2013 at 08:57:39AM -0400, Luke Faraone a écrit :
> On Sat, 2013-07-13 at 17:27 +0900, Charles Plessy wrote:
> > Le Sat, Jul 13, 2013 at 06:02:22AM +0000, Luke Faraone a écrit :
> > > 
> > > The upstream source contains pre-compiled shared object files in
> > > .\happy.hbrem\src\happy.hbrem.so and .\happy.hbrem\src-i386\happy.hbrem.so.
> > > 
> > > Please repack the tarball and reupload. While you're doing so, feel free
> > > to remove all of the ~ and #*# files that upstream left in the tarball.
> > 
> > Hi Luke and everybody,
> > 
> > It happens that I am becoming increasingly busy at work.  I have no time
> > to spend on polishing upstream's tarball cosmetically.
> 
> The REJECT reasons were not cosmetic; we cannot be distributing binaries
> for which we cannot assuredly build from source, even in source
> tarballs. 

Yes, we can distribute these binaries that are not build from source, because
they are cruft.  As you suspected, after deleting the files, everything works
fine.  To me, it demonstrates that repacking is pointless, because it provides
no service to anybody.

I can give my time for asking in a best-effort way to Upstream to clean the
cruft in the source package, but I will not waste my time repacking.

By the way, may I ask you a favor ?  We need the packages r-cran-munsell and
r-cran-scales (currently in NEW) to update the r-cran-ggplot2 package, a quite
high profile R module, which is completely outdated after the long Wheezy
freeze, and some of our users (and developers) are quite angry at us for the
mess.  Could you fast-track these two packages ?

Cheers,

-- 
Charles Plessy
Tsurumi, Kanagawa, Japan