| 1 |
On Mon, 2006-03-06 at 09:21 +0100, Kevin F. Quinn (Gentoo) wrote: |
| 2 |
> On Sun, 5 Mar 2006 20:46:25 -0500 |
| 3 |
> Mike Frysinger <vapier@g.o> wrote: |
| 4 |
> |
| 5 |
> > On Sunday 05 March 2006 19:48, Kevin F. Quinn (Gentoo) wrote: |
| 6 |
> > > This could be done via the profiles, perhaps - package.qa, something |
| 7 |
> > > like package.mask/use/keywords: |
| 8 |
> > |
| 9 |
> > i hate such things ... imo this information should stay in the ebuild |
| 10 |
> > and nowhere else ... |
| 11 |
> |
| 12 |
> I was thinking that the data would be "owned" by the QA team rather |
| 13 |
> than the package maintainers. I appreciate your pov, however. |
| 14 |
|
| 15 |
That is a good point. Sadly I'd say I do not trust a large portion of |
| 16 |
our maintainers to do the right thing and not abuse the variable in the |
| 17 |
long run simply because they do not know how to fix a pkg properly. |
| 18 |
|
| 19 |
vs say exporting env variables all over the place (bloating ebuilds |
| 20 |
even more) a single file does offer quite a few advantages. In addition |
| 21 |
a single file could be directly parsed by the scanelf util and be |
| 22 |
simply told to ignore the additional QA checks using fnmatch() for |
| 23 |
paths or so. It would also provide an easy overview of whats messed up |
| 24 |
in the tree. |
| 25 |
|
| 26 |
shrug... |
| 27 |
|
| 28 |
|
| 29 |
[snip] |
| 30 |
|
| 31 |
> Heh - here's another idea for you to hate: |
| 32 |
> |
| 33 |
> QA_OVERRIDE="EXECSTACK=... |
| 34 |
> x86? ( TEXTRELS=... )" |
| 35 |
|
| 36 |
/me hates that also. |
| 37 |
|
| 38 |
-- |
| 39 |
Ned Ludd <solar@g.o> |
| 40 |
Gentoo Linux |
| 41 |
|
| 42 |
-- |
| 43 |
gentoo-portage-dev@g.o mailing list |
Archives