1 |
On Tue, 18 Sep 2012 21:53:55 +0100 |
2 |
Ciaran McCreesh <ciaran.mccreesh@××××××××××.com> wrote: |
3 |
|
4 |
> On Tue, 18 Sep 2012 22:51:04 +0200 |
5 |
> Michał Górny <mgorny@g.o> wrote: |
6 |
> > On Tue, 18 Sep 2012 20:44:33 +0100 |
7 |
> > Ciaran McCreesh <ciaran.mccreesh@××××××××××.com> wrote: |
8 |
> > > On Tue, 18 Sep 2012 12:40:51 -0700 |
9 |
> > > Zac Medico <zmedico@g.o> wrote: |
10 |
> > > > On 09/18/2012 12:29 PM, Ciaran McCreesh wrote: |
11 |
> > > > > On Tue, 18 Sep 2012 12:25:57 -0700 |
12 |
> > > > > Zac Medico <zmedico@g.o> wrote: |
13 |
> > > > >> Also, if we change the meaning of RDEPEND in the next EAPI, |
14 |
> > > > >> so that it's a hard build-time dep like DEPEND, then |
15 |
> > > > >> DEPEND="${RDEPEND} virtual/pkgconfig" can be reduced to |
16 |
> > > > >> DEPEND="virtual/pkgconfig". This is what I would like to do |
17 |
> > > > >> for the experimental EAPI 5-hdepend which is planned [1]. |
18 |
> > > > > |
19 |
> > > > > What're we going to do about the zillions of unsolvable cycles |
20 |
> > > > > that that would create? (Does Portage detect those and error |
21 |
> > > > > out yet?) |
22 |
> > > > |
23 |
> > > > Yeah, it would be treated just like a DEPEND cycle, which is |
24 |
> > > > already detected and treated as a fatal error. As a result, when |
25 |
> > > > bumping the EAPI of an ebuild, you may have to migrate some deps |
26 |
> > > > from RDEPEND to PDEPEND in order to solve the cycles. |
27 |
> > > |
28 |
> > > What about the large number of RDEPENDs that are required for a |
29 |
> > > package to be usable, but not for it to be installed? |
30 |
> > |
31 |
> > They will still be RDEPEND, just installed earlier I believe. Except |
32 |
> > for those arising conflicts which will have to be moved to PDEP. But |
33 |
> > I think Zac said that already. |
34 |
> |
35 |
> ...but you can't move them to be a PDEPEND, since PDEPENDs aren't |
36 |
> guaranteed to be installed when a package is used. |
37 |
|
38 |
But didn't we already point out that we can't have them in RDEPEND |
39 |
since they introduce conflicts? |
40 |
|
41 |
Unless you're talking about that group of dependencies which doesn't |
42 |
introduce conflicts in RDEPEND now but will introduce them after |
43 |
the change. We should probably do some kind of tree-wide study on how |
44 |
large the problem is. |
45 |
|
46 |
A simple solution would be to mandate installing PDEPs as soon |
47 |
as possible. In case of those dependencies, that would mean installing |
48 |
them like RDEPENDs are installed now, wouldn't it? |
49 |
|
50 |
-- |
51 |
Best regards, |
52 |
Michał Górny |