| 1 |
On Sun, Sep 15, 2013 at 3:08 PM, Ciaran McCreesh |
| 2 |
<ciaran.mccreesh@××××××××××.com> wrote: |
| 3 |
> On Sun, 15 Sep 2013 11:03:28 -0400 |
| 4 |
> Rich Freeman <rich0@g.o> wrote: |
| 5 |
>> So, how about this as a policy: |
| 6 |
>> If a maintainer has an open STABLEREQ, or a KEYWORDREQ blocking a |
| 7 |
>> pending STABLEREQ, for 90 days with archs CCed and otherwise ready to |
| 8 |
>> be stabilized, the maintainer can remove older stable versions of the |
| 9 |
>> package at their discretion. A package is considered ready to be |
| 10 |
>> stabilized if it has been in the tree for 30 days, and has no known |
| 11 |
>> major flaws on arches that upstream considers supported. |
| 12 |
> |
| 13 |
> The danger of this approach is that it encourages arch teams to stable |
| 14 |
> even if they're not convinced something's been tested sufficiently. |
| 15 |
|
| 16 |
My understanding is that half the "stable" packages on the arches in |
| 17 |
question are only compile-tested anyway. If we do drop the keywords |
| 18 |
then users will only have the choice of running packages that aren't |
| 19 |
stable-tested at all. |
| 20 |
|
| 21 |
I don't really see my proposal as an ideal one - merely a compromise |
| 22 |
that seems better than the alternatives. The only thing that will |
| 23 |
actually improve the true quality of the minor arches is more |
| 24 |
manpower. If you're running s390/sparc/whatever then frankly there |
| 25 |
aren't really many better options out there, so you might as well |
| 26 |
pitch in and help scratch your own itch... |
| 27 |
|
| 28 |
Rich |
Archives