| 1 |
On Tue, 11 Sep 2012 07:58:01 -0700 |
| 2 |
Zac Medico <zmedico@g.o> wrote: |
| 3 |
> On 09/11/2012 12:33 AM, Ciaran McCreesh wrote: |
| 4 |
> > Hrm. Although... We need to decide how a :3= dep in an ebuild is |
| 5 |
> > rewritten for VDB. Do we rewrite to to :3/3.3= or to :3=3/3.3? The |
| 6 |
> > latter has the advantage that we can tell just from the VDB what the |
| 7 |
> > original spec was. |
| 8 |
> > |
| 9 |
> > The way it's worded without the patch has the benefit that stuff on |
| 10 |
> > the left of the = is written by the developer, and stuff on the |
| 11 |
> > right by the package mangler. |
| 12 |
> |
| 13 |
> Another advantage of the latter one is that when the sub-slot is |
| 14 |
> implicit, we can omit it from the rewritten dep and there's no |
| 15 |
> ambiguity about whether or not it's a rewritten dep. |
| 16 |
|
| 17 |
Wouldn't that just be rewriting to :3=3 rather than :3=3/3 then? No |
| 18 |
ambiguity either way, since the thing to the left of the = is always |
| 19 |
what's in the ebuild, and to the right is what was locked. |
| 20 |
|
| 21 |
Anyway, I think I'm retracting this patch. |
| 22 |
|
| 23 |
-- |
| 24 |
Ciaran McCreesh |
Archives