| 1 |
On 31/10/16 08:22, Raymond Jennings wrote: |
| 2 |
> I'd like to suggest that stabilization procedures be loosened a bit. |
| 3 |
> |
| 4 |
> Right now, stabilization is SLOW, and I think a lack of manpower on the |
| 5 |
> arch teams might have something to do with it. |
| 6 |
|
| 7 |
The state of stabilisation has been brought up in previous meetings and |
| 8 |
is something that is being worked on. Feel free to join us in |
| 9 |
#gentoo-wg-stable. |
| 10 |
|
| 11 |
> Suggestions: |
| 12 |
> |
| 13 |
> Have arch teams possibly certify more arch testers? If you can prove or |
| 14 |
> promise that your system is capable of testing packages for |
| 15 |
> stabilization, the arch team should be able to recruit you pretty much |
| 16 |
> no questions asked unless they have a reason otherwise. |
| 17 |
|
| 18 |
Unfortunately I've seen quite limited interest from new potential arch |
| 19 |
testers lately (compared to some years ago), and in the cases where |
| 20 |
someone did raise their hand there was nobody from the arch team |
| 21 |
available to work with them. |
| 22 |
|
| 23 |
> Also, maybe some sort of algorithm that flags the packages most badly in |
| 24 |
> need of stabilization, using the following criteria: |
| 25 |
> |
| 26 |
> * Number of days since the package or version was ~arch keyworded/added |
| 27 |
> to the tree |
| 28 |
|
| 29 |
imlate from app-portage/gentoolkit-dev can produce a report like this. |
| 30 |
|
| 31 |
> * Some sort of debianesque "popularity contest" that could (with end |
| 32 |
> user opt-in of course) flag which packages are most commonly installed |
| 33 |
> and therefore for which ones stabilization will have the greatest |
| 34 |
> positive impact for the users. |
| 35 |
|
| 36 |
gentoostats was created a long time ago but nobody actually deployed it. |
Archives