| 1 |
On Fri, 2012-07-20 at 20:17 +0100, Ciaran McCreesh wrote: |
| 2 |
> On Fri, 20 Jul 2012 15:15:31 -0400 |
| 3 |
> Alexandre Rostovtsev <tetromino@g.o> wrote: |
| 4 |
> > > That's sensitive to old versions ebuilds being removed from the |
| 5 |
> > > tree, so it's utterly unworkable. |
| 6 |
> > |
| 7 |
> > I do not see why you think it's unworkable. Ebuilds already have |
| 8 |
> > dependencies that can be broken by removing an old version; if wombat |
| 9 |
> > depends on foo[bar], and you removed the only version of foo that had |
| 10 |
> > bar in IUSE, you broke wombat. Adding special LINGUAS handling would |
| 11 |
> > not change the fact that before deleting an ebuild, you need to |
| 12 |
> > verify that you did not render other ebuilds' dependencies |
| 13 |
> > unsatisfiable. |
| 14 |
> |
| 15 |
> That's not how undefaulted use dependencies work. If wombat depends |
| 16 |
> upon foo[bar], it is an error if there is *any* version of foo *ever* |
| 17 |
> that doesn't have bar in IUSE_EFFECTIVE. |
| 18 |
|
| 19 |
Very odd; AFAICT neither portage nor repoman treats that situation as an |
| 20 |
error. I am guessing that this is another case where paludis does things |
| 21 |
differently? |
| 22 |
|
| 23 |
Be that as it may, even with paludis, the foo maintainer could easily |
| 24 |
break wombat if wombat depended on foo:bar, and the last ebuild matching |
| 25 |
foo:bar got removed; or on foo[bar,baz], and the only remaining versions |
| 26 |
of foo in the tree have REQUIRED_USE="^^ ( bar baz )"; or on foo[bar], |
| 27 |
when the only remaining versions of foo in the tree have bar disabled |
| 28 |
via profiles/base/package.use.mask. |
Archives