1 |
On Tue, Jul 25, 2017 at 9:10 AM, Michael Palimaka <kensington@g.o> wrote: |
2 |
> |
3 |
> The 30 day waiting period is useful for smoking out major upstream bugs, |
4 |
> but can't replace stabilisation integration testing. For example, |
5 |
> package foobar may build fine in ~arch but fails in stable because it |
6 |
> needs a newer libbaz. |
7 |
> |
8 |
|
9 |
I think this is a good reason why everything should be at least |
10 |
build-tested on a stable tree before getting stabilized. Now, that |
11 |
might not be on each arch if it is truly arch-independent (and if |
12 |
arches keep the dependencies reasonably in sync). |
13 |
|
14 |
This might be a situation where a compromise could exist. Have some |
15 |
kind of flag (in metadata, or maybe the ebuild) that indicates that |
16 |
the maintainer believes the package is low-risk to stabilize purely on |
17 |
a build test. Then after 30 days in testing a tinderbox could |
18 |
build-test the package and stabilize it automatically. |
19 |
|
20 |
If the package is considered at-risk then it goes through manual |
21 |
testing, either by the maintainer or an arch team. |
22 |
|
23 |
This will also encourage the teams doing testing to actually do more |
24 |
testing on the packages that would benefit from it. |
25 |
|
26 |
We wouldn't set hard criteria but leave it up to maintainer |
27 |
discretion, with a guideline being that past performance is probably a |
28 |
good predictor of future results. |
29 |
|
30 |
-- |
31 |
Rich |