1 |
El vie, 08-06-2012 a las 12:16 -0700, Zac Medico escribió: |
2 |
> On 06/08/2012 01:38 AM, Pacho Ramos wrote: |
3 |
> > El jue, 07-06-2012 a las 12:33 -0700, Zac Medico escribió: |
4 |
> >> On 06/07/2012 12:24 PM, Pacho Ramos wrote: |
5 |
> >>> El jue, 07-06-2012 a las 12:09 -0700, Zac Medico escribió: |
6 |
> >>>> On 06/07/2012 12:00 PM, Pacho Ramos wrote: |
7 |
> >>>>> El jue, 07-06-2012 a las 19:44 +0100, Ciaran McCreesh escribió: |
8 |
> >>>>>> On Thu, 07 Jun 2012 20:43:54 +0200 |
9 |
> >>>>>> Pacho Ramos <pacho@g.o> wrote: |
10 |
> >>>>>>>> I would prefer, as a workaround, allow reverse deps to RDEPEND on |
11 |
> >>>>>>>> glib:2.* instead. That way it would cover more cases when more than |
12 |
> >>>>>>>> two slots are available |
13 |
> >>>>>>> |
14 |
> >>>>>>> Well, per: |
15 |
> >>>>>>> http://git.overlays.gentoo.org/gitweb/?p=proj/pms.git;a=commitdiff;h=f9f7729c047300e1924ad768a49c660e12c2f906;hp=b7750e67b4772c1064543defb7df6a556f09807b |
16 |
> >>>>>>> |
17 |
> >>>>>>> looks like "*" usage for SLOTs would be allowed :), or I am |
18 |
> >>>>>>> misinterpreting it? |
19 |
> >>>>>> |
20 |
> >>>>>> It's not a wildcard. |
21 |
> >>>>>> |
22 |
> >>>>> |
23 |
> >>>>> But it looks like a valid usage for cases like glib vs. |
24 |
> >>>>> dbus-glib/gobject-introspection I have exposed as example, and also |
25 |
> >>>>> allows us to keep "SLOT" over "ABI_SLOT" (at least for this case, not |
26 |
> >>>>> sure about others I could be missing now...) |
27 |
> >>>> |
28 |
> >>>> The :* operator doesn't trigger any rebuilds though. Quoting the PMS |
29 |
> >>>> patch that you linked: |
30 |
> >>>> |
31 |
> >>>> * Indicates that any slot value is acceptable. In addition, for runtime |
32 |
> >>>> dependencies, indicates that the package will not break if the matched |
33 |
> >>>> package is uninstalled and replaced by a different matching package in a |
34 |
> >>>> different slot. |
35 |
> >>> |
36 |
> >>> I mean, use it in conjunction with ":=", one for rebuild and other to |
37 |
> >>> indicate any 2.x SLOT fits the "normal" RDEPEND (to not need to |
38 |
> >>> periodically update RDEPENDs or need to go back from :SLOT depends to |
39 |
> >>> old =category/package-version-* ways) |
40 |
> >>> |
41 |
> >>> Allowing that, we wouldn't need ABI_SLOT (at least to prevent this issue |
42 |
> >>> that arises with using only SLOTs for this) |
43 |
> >> |
44 |
> >> What you're talking about here is more similar to ABI_SLOT operator deps |
45 |
> >> than what was originally intended for SLOT operator deps. In other |
46 |
> >> words, anyone who is opposed to ABI_SLOT operator deps is likely to also |
47 |
> >> be opposed to your proposal. |
48 |
> > |
49 |
> > Oh :(, and what is the reason to want to prevent this behavior? Looks |
50 |
> > much simpler to me than needing to use ranges for dependencies or |
51 |
> > needing to create "compat" packages to hide the problem :| |
52 |
> |
53 |
> It's close enough to ABI_SLOT that it would make more sense just to use |
54 |
> ABI_SLOT because it's more flexible. |
55 |
|
56 |
In that case, I think it's clear we need ABI_SLOT ;) The problem is how |
57 |
to document it in a way people agree with including it for eapi5 :| |