1 |
On Tue, 18 Sep 2012 12:58:30 -0700 |
2 |
Zac Medico <zmedico@g.o> wrote: |
3 |
> On 09/18/2012 12:44 PM, Ciaran McCreesh wrote: |
4 |
> > On Tue, 18 Sep 2012 12:40:51 -0700 |
5 |
> > Zac Medico <zmedico@g.o> wrote: |
6 |
> >> On 09/18/2012 12:29 PM, Ciaran McCreesh wrote: |
7 |
> >>> On Tue, 18 Sep 2012 12:25:57 -0700 |
8 |
> >>> Zac Medico <zmedico@g.o> wrote: |
9 |
> >>>> Also, if we change the meaning of RDEPEND in the next EAPI, so |
10 |
> >>>> that it's a hard build-time dep like DEPEND, then |
11 |
> >>>> DEPEND="${RDEPEND} virtual/pkgconfig" can be reduced to |
12 |
> >>>> DEPEND="virtual/pkgconfig". This is what I would like to do for |
13 |
> >>>> the experimental EAPI 5-hdepend which is planned [1]. |
14 |
> >>> |
15 |
> >>> What're we going to do about the zillions of unsolvable cycles |
16 |
> >>> that that would create? (Does Portage detect those and error out |
17 |
> >>> yet?) |
18 |
> >> |
19 |
> >> Yeah, it would be treated just like a DEPEND cycle, which is |
20 |
> >> already detected and treated as a fatal error. As a result, when |
21 |
> >> bumping the EAPI of an ebuild, you may have to migrate some deps |
22 |
> >> from RDEPEND to PDEPEND in order to solve the cycles. |
23 |
> > |
24 |
> > What about the large number of RDEPENDs that are required for a |
25 |
> > package to be usable, but not for it to be installed? |
26 |
> |
27 |
> You will have to migrate those deps from RDEPEND to PDEPEND. |
28 |
|
29 |
...but PDEPENDs aren't guaranteed to be installed before a package is |
30 |
used to satisfy a dependency. (And we can't change PDEPEND to do what |
31 |
RDEPEND currently does, because then some cycles can't be solved at |
32 |
all.) |
33 |
|
34 |
-- |
35 |
Ciaran McCreesh |