| 1 |
On Sun, Jun 27, 2010 at 8:34 PM, Markos Chandras <hwoarang@g.o> wrote: |
| 2 |
> As many of you have already noticed, there are some arches that are quite |
| 3 |
> slow on stabilizations. This leads to deprecated stabilizations e.g a |
| 4 |
> package is stabilized after 60 days which makes that version of |
| 5 |
> the specific package obsolete and not worth to stabilize anymore. |
| 6 |
> |
| 7 |
> I would suggest to introduce a new rule where a stabilization bug may close |
| 8 |
> after 30 days. Arches that fail to stabilize it within this timeframe |
| 9 |
> they will simply don't have this package stable for them. |
| 10 |
> |
| 11 |
|
| 12 |
I disagree that just 30 days is enough to make a package obsolete and |
| 13 |
not worth stabilizing. Infact, I think you're suggesting the number in |
| 14 |
jest. GNOME for instance has a 6 month cycle, and we stabilize 1-4 |
| 15 |
months after the release (depending on the bugs, and our manpower). |
| 16 |
For instance, the current release is 2.30 (~arch, released in March |
| 17 |
'10, added to tree just a month ago), the previous release was 2.28 |
| 18 |
(stable, released in Sept '09), but the stable version on every arch |
| 19 |
except amd64/x86 is 2.26 (released in March '09). And that's OK since |
| 20 |
it got stable updates for a long time after release, and works |
| 21 |
wonderfully. In essence, GNOME packages don't become obsolete unless |
| 22 |
it's been 2 years since they were released. |
| 23 |
|
| 24 |
I'm saying that a 30 days rule is too strict for most packages and |
| 25 |
herds. I don't think such a rule will fly very far. Even a 90 day rule |
| 26 |
or a 6 month rule is too strict for GNOME packages. I personally |
| 27 |
empathize with the needs of users enough that I (and most of the gnome |
| 28 |
team) are willing to wait for arches that cannot handle stabilization |
| 29 |
bugs. We really don't want our users to have a bad experience because |
| 30 |
of *us*. We'll do whatever is in our power. |
| 31 |
|
| 32 |
|
| 33 |
|
| 34 |
> Moreover, slow arches introduce another problem as well. If a package is |
| 35 |
> marked stabled for their arch, but this package is quite old, and they fail to |
| 36 |
> stabilize a new version, we ( as maintainers ) can't drop the very old |
| 37 |
> ( and obsolete ) version of this package because we somehow will break |
| 38 |
> the stable tree for these arches. How should we act in this case? |
| 39 |
> Keep the old version around forever just to say that "hey, they do have |
| 40 |
> a stable version for our exotic arch". |
| 41 |
> |
| 42 |
|
| 43 |
Now *this* is a problem. We have some bugs, some security bugs that |
| 44 |
have been completely ignored by some arches. Mips as usual is one, but |
| 45 |
recently hppa (and to a much lesser extent, ppc64) have become slow. |
| 46 |
|
| 47 |
To fix this, I suggest the following heuristic: |
| 48 |
|
| 49 |
* If an arch cannot stabilize *security bugs* after 3 months, the |
| 50 |
maintainers are free to drop the vulnerable version. |
| 51 |
* If an arch cannot stabilize versions which fix major bugs after 6 |
| 52 |
months, the maintainers are free to drop the broken version. |
| 53 |
* If an arch cannot stabilize a feature/minor bugfix stable request |
| 54 |
after 12 months, the maintainers can nuke stable keywords from their |
| 55 |
package. |
| 56 |
|
| 57 |
In all cases, the time is to be counted from the original |
| 58 |
stabilization request. Exceptions may be made in case newer |
| 59 |
stabilization requests add extra dependencies to be stabilized. |
| 60 |
|
| 61 |
Similar (but laxer) rules can be made for KEYWORDREQ bugs as well. |
| 62 |
|
| 63 |
-- |
| 64 |
~Nirbheek Chauhan |
| 65 |
|
| 66 |
Gentoo GNOME+Mozilla Team |
Archives