| 1 |
Zac Medico wrote: |
| 2 |
> -----BEGIN PGP SIGNED MESSAGE----- |
| 3 |
> Hash: SHA1 |
| 4 |
> |
| 5 |
> Daniel Barkalow wrote: |
| 6 |
>> If the configuration has keywords "foo bar", and a package has "-foo |
| 7 |
>> bar", mask the package ("masked by -bar keyword"). |
| 8 |
>> |
| 9 |
>> This is the sensible behavior if we ever make use of listing multiple |
| 10 |
>> keywords in the configuration, which is currently implemented but not |
| 11 |
>> used for anything. |
| 12 |
> |
| 13 |
> Personally, I'd prefer not to support -foo or -* in the KEYWORDS of |
| 14 |
> an ebuild. For one thing, seems like it's trying to accomplish |
| 15 |
> something similar to what package.mask is intended for. Another |
| 16 |
> problem is that is uses -foo and -* in completely different ways |
| 17 |
> than they are used elsewhere in portage (for negation of values in |
| 18 |
> an incremental stack). |
| 19 |
> |
| 20 |
> Zac |
| 21 |
|
| 22 |
|
| 23 |
Didn't someone write a profile introspection tool to look at stuff like |
| 24 |
this? If it was as simple as emask profile package, and the tool took |
| 25 |
care of many of the details I think there would be less argument. As it |
| 26 |
is, setting -foo in the ebuild is trivial compared to the alternative |
| 27 |
(since profiles stack and you may need to edit 2 or 3 in some cases, |
| 28 |
means greater chance of screwing up). |
| 29 |
-- |
| 30 |
gentoo-portage-dev@g.o mailing list |
Archives