| 1 |
On Thu, 2006-07-27 at 09:24 -0600, Steve Dibb wrote: |
| 2 |
> Chris Gianelloni wrote: |
| 3 |
> >> I'd say no bugs, 30 days, passes internal tests, being run by users => |
| 4 |
> >> stablise, for the majority of packages (obviously, there may be some |
| 5 |
> >> exceptions...). |
| 6 |
> >> |
| 7 |
> > |
| 8 |
> > Luckily, you're not making the call. ;] |
| 9 |
> > |
| 10 |
> > The "majority" of packages are also the ones that need more extensive |
| 11 |
> > testing. Sure, we could probably stabilize a bunch of the fringe |
| 12 |
> > packages that hardly anyone uses and it wouldn't affect anything. |
| 13 |
> |
| 14 |
> That's actually how I read the first email, was that it's really the |
| 15 |
> majority of the _minor_ packages that get completely neglected, and just |
| 16 |
> sits in the tree for months or years marked unstable because nobody |
| 17 |
> cares. The people that use it have marked it ~arch a long time ago in |
| 18 |
> their package.keywords because they know it works just fine. |
| 19 |
|
| 20 |
Well, we would hope that people using the package would file a bug, but |
| 21 |
this obviously doesn't always happen. |
| 22 |
|
| 23 |
> THAT stuff I wouldn't mind going through and just bumping to stable |
| 24 |
> myself. They don't need extensive testing, they don't need patches, |
| 25 |
> they work, and have been working, and just need arches flagged and |
| 26 |
> versions bumped. |
| 27 |
|
| 28 |
I'd have no problem with that so long as it was done by a person. I |
| 29 |
just don't trust that anything like this should ever be automated. |
| 30 |
|
| 31 |
Now, we could have an automated system to send out *alerts* on packages |
| 32 |
that have been in testing for more than 30 (or 60) days. We could even |
| 33 |
make it searchable by architecture and put it on a web page. We could |
| 34 |
probably also make it give some information about QA problems. We could |
| 35 |
even make it searchable by herd. That would be cool. |
| 36 |
|
| 37 |
I propose that we put up such a page and we call it |
| 38 |
http://gentoo.tamperd.net/stable/ and make it publicly available. ;] |
| 39 |
|
| 40 |
> But, nobody likes doing the small stuff, and I can't blame them. |
| 41 |
|
| 42 |
True that. |
| 43 |
|
| 44 |
-- |
| 45 |
Chris Gianelloni |
| 46 |
Release Engineering - Strategic Lead |
| 47 |
x86 Architecture Team |
| 48 |
Games - Developer |
| 49 |
Gentoo Linux |
Archives