1 |
В Сбт, 05/01/2008 в 18:17 -0600, Ryan Hill пишет: |
2 |
> I don't know, I can kinda see both sides. Alt arches tend to be finicky |
3 |
> so it's important that updates are well tested on them. Also they're |
4 |
> more prone to break during upgrades, not only because they're more |
5 |
> fragile but because upstream is far less likely to have tested on them, |
6 |
> so I can see why having a stable tree is important. |
7 |
|
8 |
If this is an issue arch developers should tell us that. |
9 |
|
10 |
Problems with having slacker archs are: open bugs in bugzilla and old |
11 |
ebuilds which are unsupported by maintainer[1]. Open bug just takes my |
12 |
time and attention to open it and to find out that we already fixed that |
13 |
bug and wait for arch to take their action. Old ebuilds they leave me |
14 |
without satisfaction and lie to our users - I know that they are broken, |
15 |
but they are still in the tree and are marked as stable. |
16 |
|
17 |
Open bugs problem can't be solved until we fix problem with old ebuilds |
18 |
because ordinary for broken/old ebuilds I keep herd/myself in CC of bug |
19 |
until it's closed to drop old ebuild from the tree. |
20 |
|
21 |
And for me the problem with old ebuilds could be solved if I could drop |
22 |
keywords from old ebuilds. Then I could remove herd/myself from CC to |
23 |
bug. Also if council decide this way I'd like to see recommendation for |
24 |
slacker arch to drop old ebuild (with none keywords except ~arch) from |
25 |
the tree by themselves as soon as they stabilize new version. |
26 |
|
27 |
[1] And security problem could be solved by labeling arch as security |
28 |
unsupported. |
29 |
|
30 |
-- |
31 |
Peter. |