| 1 |
On 12/05/14 05:59, Pacho Ramos wrote: |
| 2 |
> Hi! |
| 3 |
> |
| 4 |
> We found out that pulseaudio ebuild was modified by QA without QA |
| 5 |
> talking to the maintainers (gnome team) and without considering/updating |
| 6 |
> the relevant bugzilla issue at |
| 7 |
> https://bugs.gentoo.org/show_bug.cgi?id=519530 |
| 8 |
> |
| 9 |
> In that link it's explained a bit more why the ebuild was written in |
| 10 |
> that way and the problems we try to avoid. We have then hardmasked that |
| 11 |
> version until it's discussed THERE how to handle that situations. |
| 12 |
> |
| 13 |
> And would really appreciate that next time we are even notified about a |
| 14 |
> change is going to be committed and don't need to see it in |
| 15 |
> packages.gentoo.org (well, in my case I am not all the time on IRC... |
| 16 |
> but I read the mail often and, also, Gilles and Leio can also be |
| 17 |
> contacted on IRC. You can also simply send a mail to the alias and give |
| 18 |
> us at least some days of timeout). |
| 19 |
> |
| 20 |
> Thanks a lot |
| 21 |
> |
| 22 |
> |
| 23 |
|
| 24 |
I don't know the policy (I will read the relevant docs later) but its |
| 25 |
seems to me to make good sense that, if it is not an emergency (ie the |
| 26 |
tree is broken), that QA first inform the maintainer in a bug report |
| 27 |
which can then be peer reviewed. QA can make mistakes (as in this case) |
| 28 |
and that's okay if there is discussion. If it is an emergency, then I |
| 29 |
would think QA should take the action of least interference to unbreak |
| 30 |
the tree. |
| 31 |
|
| 32 |
Bikeshed time ... |
| 33 |
|
| 34 |
-- |
| 35 |
Anthony G. Basile, Ph.D. |
| 36 |
Gentoo Linux Developer [Hardened] |
| 37 |
E-Mail : blueness@g.o |
| 38 |
GnuPG FP : 1FED FAD9 D82C 52A5 3BAB DC79 9384 FA6E F52D 4BBA |
| 39 |
GnuPG ID : F52D4BBA |
Archives