| 1 |
On Sunday 06 August 2006 00:26, Mike Frysinger wrote: |
| 2 |
> and i'm on the opposite side where implicit RDEPEND should be clean: |
| 3 |
|
| 4 |
Why? I for one consider explicit dependencies much more clean. If Portage at |
| 5 |
some point should distinct between dependencies defined in ebuilds and |
| 6 |
eclasses, we'd need a defined way to set eclass dependencies in ebuilds, so |
| 7 |
Portage actually can do the distiction, not breaking the tree. |
| 8 |
|
| 9 |
> - eclass and ebuilds have their own sets of DEPEND/RDEPEND which do not in |
| 10 |
> any way affect each other |
| 11 |
|
| 12 |
That's not true. We use and need the functionality to set dependencies in the |
| 13 |
ebuild which take effect in the eclass. Be it by setting a variable before |
| 14 |
inherit or by an eclass function called from within the ebuild - need-kde(), |
| 15 |
need-apache(), ... |
| 16 |
|
| 17 |
We can't source the eclass and have all its dependencies. |
| 18 |
|
| 19 |
|
| 20 |
Carsten |
Archives