On Fri, Jan 13, 2012 at 01:12:32PM -0800, Russ Allbery wrote:
>
> Do we care? Most upstream distribution tarballs are likewise going to
> feature random UIDs and GIDs in the tarball because that's what you get
> from tar if you don't take any special precautions.
I admit that the issue might not be very important but I personally like
reproducible results (in the sense of same MD5sums). At least it can
not harm even if I admit that upstream does not care - in some points we
are better than upstream. So why not doing it if it comes cheap?
Finally the author of uscan seems to have tried to code in this sense
but did not completely succeeded.
> All of our software
> therefore already copes with this.
That's correct and thus I would use severity wishlist or minor in case
it does not turn out as a completely stupid idea.
> The tarball already isn't reproducible without tools like pristine-tar due
> to other issues with the format.
Well, pristine-tar will not help us here (or am I missing something) because
I'm talking about *re*packaging upstream (for whatever reason). Could you
please be more verbose about "other issues with the format"?
Kind regards
Andreas.
--
http://fam-tille.de
--
To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/20120113212129.gf20...@an3as.eu
