| 1 |
On Wed, Apr 2, 2014 at 3:55 PM, Mike Gilbert <floppym@g.o> wrote: |
| 2 |
> On the packages I maintain, I tend to use the latest unstable version |
| 3 |
> of the software. Stabilizing them rarely crosses my mind. |
| 4 |
> |
| 5 |
> I rather like the semi-automated reminders. They come in handy for my |
| 6 |
> own packages, as well as the large, uncoordinated mess of libraries |
| 7 |
> that the python team is tasked with maintaining. |
| 8 |
> |
| 9 |
|
| 10 |
++ |
| 11 |
|
| 12 |
I see no harm in people filing STABLEREQs when software has been |
| 13 |
around for a while. If it is inappropriate to stabilize a package |
| 14 |
without further coordination, just close the bug INVALID or such with |
| 15 |
a brief explanation. If a non-dev contributor wants to do the |
| 16 |
necessary coordination to file a better bug more power to them, and if |
| 17 |
not it is up to the maintainer. |
| 18 |
|
| 19 |
The bugs should be going to the maintainer first, and they should be |
| 20 |
responding one way or another in any case. If Arch teams are getting |
| 21 |
bugged with invalid STABLEREQs I'd say the problem is that the package |
| 22 |
should be maintainer-wanted. |
| 23 |
|
| 24 |
Another option might be to have a tag in metadata.xml that flags |
| 25 |
packages as never-stable or indicating that stabilization requires |
| 26 |
coordination, which might help with triage or getting bug filters to |
| 27 |
check first. |
| 28 |
|
| 29 |
Rich |
Archives