1 |
On 06/08/2012 12:23 PM, Pacho Ramos wrote: |
2 |
> El vie, 08-06-2012 a las 12:16 -0700, Zac Medico escribió: |
3 |
>> On 06/08/2012 01:38 AM, Pacho Ramos wrote: |
4 |
>>> El jue, 07-06-2012 a las 12:33 -0700, Zac Medico escribió: |
5 |
>>>> On 06/07/2012 12:24 PM, Pacho Ramos wrote: |
6 |
>>>>> El jue, 07-06-2012 a las 12:09 -0700, Zac Medico escribió: |
7 |
>>>>>> On 06/07/2012 12:00 PM, Pacho Ramos wrote: |
8 |
>>>>>>> El jue, 07-06-2012 a las 19:44 +0100, Ciaran McCreesh escribió: |
9 |
>>>>>>>> On Thu, 07 Jun 2012 20:43:54 +0200 |
10 |
>>>>>>>> Pacho Ramos <pacho@g.o> wrote: |
11 |
>>>>>>>>>> I would prefer, as a workaround, allow reverse deps to RDEPEND on |
12 |
>>>>>>>>>> glib:2.* instead. That way it would cover more cases when more than |
13 |
>>>>>>>>>> two slots are available |
14 |
>>>>>>>>> |
15 |
>>>>>>>>> Well, per: |
16 |
>>>>>>>>> http://git.overlays.gentoo.org/gitweb/?p=proj/pms.git;a=commitdiff;h=f9f7729c047300e1924ad768a49c660e12c2f906;hp=b7750e67b4772c1064543defb7df6a556f09807b |
17 |
>>>>>>>>> |
18 |
>>>>>>>>> looks like "*" usage for SLOTs would be allowed :), or I am |
19 |
>>>>>>>>> misinterpreting it? |
20 |
>>>>>>>> |
21 |
>>>>>>>> It's not a wildcard. |
22 |
>>>>>>>> |
23 |
>>>>>>> |
24 |
>>>>>>> But it looks like a valid usage for cases like glib vs. |
25 |
>>>>>>> dbus-glib/gobject-introspection I have exposed as example, and also |
26 |
>>>>>>> allows us to keep "SLOT" over "ABI_SLOT" (at least for this case, not |
27 |
>>>>>>> sure about others I could be missing now...) |
28 |
>>>>>> |
29 |
>>>>>> The :* operator doesn't trigger any rebuilds though. Quoting the PMS |
30 |
>>>>>> patch that you linked: |
31 |
>>>>>> |
32 |
>>>>>> * Indicates that any slot value is acceptable. In addition, for runtime |
33 |
>>>>>> dependencies, indicates that the package will not break if the matched |
34 |
>>>>>> package is uninstalled and replaced by a different matching package in a |
35 |
>>>>>> different slot. |
36 |
>>>>> |
37 |
>>>>> I mean, use it in conjunction with ":=", one for rebuild and other to |
38 |
>>>>> indicate any 2.x SLOT fits the "normal" RDEPEND (to not need to |
39 |
>>>>> periodically update RDEPENDs or need to go back from :SLOT depends to |
40 |
>>>>> old =category/package-version-* ways) |
41 |
>>>>> |
42 |
>>>>> Allowing that, we wouldn't need ABI_SLOT (at least to prevent this issue |
43 |
>>>>> that arises with using only SLOTs for this) |
44 |
>>>> |
45 |
>>>> What you're talking about here is more similar to ABI_SLOT operator deps |
46 |
>>>> than what was originally intended for SLOT operator deps. In other |
47 |
>>>> words, anyone who is opposed to ABI_SLOT operator deps is likely to also |
48 |
>>>> be opposed to your proposal. |
49 |
>>> |
50 |
>>> Oh :(, and what is the reason to want to prevent this behavior? Looks |
51 |
>>> much simpler to me than needing to use ranges for dependencies or |
52 |
>>> needing to create "compat" packages to hide the problem :| |
53 |
>> |
54 |
>> It's close enough to ABI_SLOT that it would make more sense just to use |
55 |
>> ABI_SLOT because it's more flexible. |
56 |
> |
57 |
> In that case, I think it's clear we need ABI_SLOT ;) The problem is how |
58 |
> to document it in a way people agree with including it for eapi5 :| |
59 |
|
60 |
We can just write a specification for this one feature, and ask the |
61 |
Council to approve it. |
62 |
-- |
63 |
Thanks, |
64 |
Zac |