1 |
On Mon, 2014-01-27 at 09:52 -0500, Rich Freeman wrote: |
2 |
> On Mon, Jan 27, 2014 at 2:41 AM, Steev Klimaszewski <steev@g.o> wrote: |
3 |
> > It's not necessarily the STABLEREQs stopping, some of the issues are (at |
4 |
> > least on some arches!) that some of the unstable software doesn't quite |
5 |
> > work properly anymore, and we are failing at communicating. And in |
6 |
> > those cases, we on the arch teams should definitely be pointing this |
7 |
> > out, and filing bugs so that the issues can be sorted. |
8 |
> |
9 |
> Well, if the package or some version of it doesn't work at all, you |
10 |
> can always mask it on the arch or drop keywords. The arch team |
11 |
> doesn't need permission to do this stuff - the keywords and profiles |
12 |
> really "belong" to the arch team, and we just allow maintainers to do |
13 |
> their best job with them to make the job of the arch team easier. |
14 |
> |
15 |
|
16 |
Right, but, afaik, an "unstable" ebuild can go away at any point in |
17 |
time, and then we'd be back in this same place - newer ebuilds are |
18 |
around, older working ones are gone... |
19 |
|
20 |
> Obviously if you actually want the problem fixed that requires |
21 |
> bugs/etc. But you don't need a bug to drop a keyword and at least |
22 |
> make it clear that the package doesn't work. |
23 |
> |
24 |
|
25 |
Right, and this goes as a point towards splitting out the arm keywords, |
26 |
and maybe I'll bring it up at the next ARM team meeting... I don't think |
27 |
it would get much traction, but I suppose it wouldn't hurt to at least |
28 |
throw it out there and see what sticks. |
29 |
|
30 |
> Rich |
31 |
> |