Toshio Kuratomi schrieb:
I'm arguing that the problem Karsten thinks he's addressing is
"widespread use of the autotools when there's no need to do it". If
there are few packages that BuildRequire autotools in the first place,
then there can't be a widespread use of autotools in package building,
let alone a widespread use that is unnecessary.
-Toshio
I think you misunderstood my proposal to drop the old autofoo stuff.
The main reasoning for my proposal was that I think that software packages shouldn't rely
on old, unmaintained (upstream) helper programs such as p.e. automake-1.4 and that we
should help upstream to move to more recent autofoo. Raising a barrier by not shipping the
old stuff anymore and thus maybe forcing upstream maintainers to other distributions was a
bad idea which I've canceled during this discussion. But I still think a guideline that
new packages should be checked if they can easily ported to current autofoo before they
get accepted would help us und upstream in the longer term. Please note that I don't insist
on having them ported, if it is too complicated to port it should still get accepted.
But not every package has that many special cases and hacks as the firefox package, most
should be portable without too much affort and I'm sure most upstream maintainers would be
glad to get patches for the autofoo stuff.
Karsten
--
fedora-devel-list mailing list
fedora-devel-list@xxxxxxxxxx
https://www.redhat.com/mailman/listinfo/fedora-devel-list
