| 1 |
On Tue, 2002-12-17 at 12:40, Sean P. Kane wrote: |
| 2 |
> All this |
| 3 |
> does is point out the real lesson to be learned here. Gentoo may be |
| 4 |
> "cutting-edge", but the stability problems MUST be fixed, expecially in |
| 5 |
> Portafge, because if it breaks, it can leave a system in a very |
| 6 |
> difficult to recover state. |
| 7 |
|
| 8 |
We actually have a good QA procedure in place for Portage. The problem |
| 9 |
was caused by a senior developer doing things that he shouldn't and not |
| 10 |
following this procedure. |
| 11 |
|
| 12 |
Again, I don't want to make this person feel worse than he already does |
| 13 |
(and he does feel bad about his error in judgement.) I sharing this info |
| 14 |
for the sole purpose of letting everyone know that normally this problem |
| 15 |
would have been caught way before it hit end-users, and that we *do* |
| 16 |
have a QA procedure in place for Portage. |
| 17 |
|
| 18 |
Today, I've made extra sure that all our devs are very clear on the |
| 19 |
rules regarding new Portage releases. I've done what's necessary to help |
| 20 |
ensure that issues like this are avoided in the future. I am now going |
| 21 |
to incorporate these details regarding Portage QA into our official |
| 22 |
policy documentation on the Web site so that future developers will do |
| 23 |
the right thing too. |
| 24 |
|
| 25 |
Sincerely, |
| 26 |
|
| 27 |
-- |
| 28 |
Daniel Robbins |
| 29 |
Chief Architect, Gentoo Linux |
| 30 |
http://www.gentoo.org |
Archives