| 1 |
>>>>> On Mon, 07 May 2012, Zac Medico wrote: |
| 2 |
|
| 3 |
> Another possible way to model this kind of relationship would be to |
| 4 |
> us REQUIRED_USE to enforce relationships with the qt and gtk flags: |
| 5 |
|
| 6 |
> REQUIRED_USE="webkit? ( qt ) !webkit? ( !qt ) qt? ( webkit ) !qt? ( |
| 7 |
> !webkit )" |
| 8 |
|
| 9 |
This line just says that either both webkit and qt must be set, or |
| 10 |
neither of them. In other words, that the ebuild has redundant flags, |
| 11 |
which IMHO should be avoided. |
| 12 |
|
| 13 |
> versus |
| 14 |
|
| 15 |
> REQUIRED_USE="webkit? ( gtk ) !webkit? ( !gtk ) gtk? ( webkit ) !gtk? ( |
| 16 |
> !webkit )" |
| 17 |
|
| 18 |
> It's pretty awkward with the existing operators, but we could extend |
| 19 |
> the REQUIRED_USE syntax to support an equivalent operator in a |
| 20 |
> future EAPI. |
| 21 |
|
| 22 |
As far as I can see, it is equivalent to: |
| 23 |
|
| 24 |
REQUIRED_USE="^^ ( webkit !qt )" |
| 25 |
|
| 26 |
(or analog for the gtk case). But see above. |
| 27 |
|
| 28 |
Ulrich |
Archives