| 1 |
On Wed, 2008-01-09 at 18:56 +0100, Jan Kundrát wrote: |
| 2 |
> Chris Gianelloni wrote: |
| 3 |
> > I have foo 1.0, which is mips. There is foo 2.0, which is stable |
| 4 |
> > everywhere else. The foo 1.0 ebuild does not conform to current ebuild |
| 5 |
> > standards. I want to commit changes to foo 2.0, and repoman won't allow |
| 6 |
> > me due to problems in foo 1.0, but I don't want to WASTE MY TIME on foo |
| 7 |
> > 1.0, because it's been EOL for 2 years |
| 8 |
> |
| 9 |
> Why don't fix repoman not to scream about such issues, then? |
| 10 |
|
| 11 |
What, have repoman complain only about problems in ebuilds that have |
| 12 |
been changed unless someone does "repoman full" ? |
| 13 |
|
| 14 |
Honestly, that coupled with dropping all KEYWORDS except for the arch in |
| 15 |
question (in other words, marking something KEYWORDS="mips" and then |
| 16 |
ignoring it, as a maintainer) would be enough to keep package |
| 17 |
maintainers and other architecture teams from having to deal with the |
| 18 |
crap left all over the tree due to slacker arches. Of course, tree |
| 19 |
quality would probably go down even more, since these QA issues would |
| 20 |
likely never be fixed on said architectures, but who really cares, |
| 21 |
anyway. The support burden gets lain on the people who are slacking, |
| 22 |
and not on the package maintainers or other architecture teams. |
| 23 |
|
| 24 |
-- |
| 25 |
Chris Gianelloni |
| 26 |
Release Engineering Strategic Lead |
| 27 |
Alpha/AMD64/x86 Architecture Teams |
| 28 |
Games Developer |
Archives