* [gentoo-pms] Clarification of has_version cat/pkg[use=] behavior for EAPI 3 or 4
@ 2009-09-12 17:41 Petteri Räty
2009-09-12 18:11 ` Ciaran McCreesh
2009-09-12 19:43 ` Christian Faulhammer
0 siblings, 2 replies; 7+ messages in thread
From: Petteri Räty @ 2009-09-12 17:41 UTC (permalink / raw
To: gentoo-pms
Currently Portage does not work with has_version cat/pkg[use=]. I filed
a bug about this:
https://bugs.gentoo.org/show_bug.cgi?id=284715
Portage current behavior:
17:38 <@zmedico> Betelgeuse: quick test seems to show always true if
it's in IUSE
I see a couple course of possible actions:
1. Current PMS already implies this behavior and Portage is buggy. Let's
document this explicitly for EAPI 3 and have Portage fixed so using EAPI
3 makes sure this syntax works.
2. PMS currently does not specify this properly and as such this would
belong to EAPI 4.
3. Suggestions welcome.
Regards,
Petteri
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [gentoo-pms] Clarification of has_version cat/pkg[use=] behavior for EAPI 3 or 4
2009-09-12 17:41 [gentoo-pms] Clarification of has_version cat/pkg[use=] behavior for EAPI 3 or 4 Petteri Räty
@ 2009-09-12 18:11 ` Ciaran McCreesh
2009-09-12 18:13 ` Petteri Räty
2009-09-12 19:43 ` Christian Faulhammer
1 sibling, 1 reply; 7+ messages in thread
From: Ciaran McCreesh @ 2009-09-12 18:11 UTC (permalink / raw
To: gentoo-pms
[-- Attachment #1: Type: text/plain, Size: 1207 bytes --]
On Sat, 12 Sep 2009 20:41:59 +0300
Petteri Räty <betelgeuse@gentoo.org> wrote:
> Currently Portage does not work with has_version cat/pkg[use=]. I
> filed a bug about this:
> https://bugs.gentoo.org/show_bug.cgi?id=284715
>
> Portage current behavior:
> 17:38 <@zmedico> Betelgeuse: quick test seems to show always true if
> it's in IUSE
>
> I see a couple course of possible actions:
> 1. Current PMS already implies this behavior and Portage is buggy.
> Let's document this explicitly for EAPI 3 and have Portage fixed so
> using EAPI 3 makes sure this syntax works.
The wording is "The flag must be enabled if the flag is enabled for the
package with the dependency, or disabled otherwise.".
The question then is whether that applies only to *DEPEND, or whether
it's to anywhere where there's a "package with the dependency". Given
that the section that specifies it isn't limited to *DEPEND, but
describes dependency specifications in general, I'd say that PMS
requires it to be supported.
Having said that, it's pretty horrible, and unless there's a good use
case for such a feature, I'd not be opposed to shoving in a retroactive
ban on it...
--
Ciaran McCreesh
[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 198 bytes --]
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [gentoo-pms] Clarification of has_version cat/pkg[use=] behavior for EAPI 3 or 4
2009-09-12 18:11 ` Ciaran McCreesh
@ 2009-09-12 18:13 ` Petteri Räty
2009-09-12 18:20 ` Ciaran McCreesh
0 siblings, 1 reply; 7+ messages in thread
From: Petteri Räty @ 2009-09-12 18:13 UTC (permalink / raw
To: gentoo-pms
Ciaran McCreesh wrote:
> On Sat, 12 Sep 2009 20:41:59 +0300
> Petteri Räty <betelgeuse@gentoo.org> wrote:
>> Currently Portage does not work with has_version cat/pkg[use=]. I
>> filed a bug about this:
>> https://bugs.gentoo.org/show_bug.cgi?id=284715
>>
>> Portage current behavior:
>> 17:38 <@zmedico> Betelgeuse: quick test seems to show always true if
>> it's in IUSE
>>
>> I see a couple course of possible actions:
>> 1. Current PMS already implies this behavior and Portage is buggy.
>> Let's document this explicitly for EAPI 3 and have Portage fixed so
>> using EAPI 3 makes sure this syntax works.
>
> The wording is "The flag must be enabled if the flag is enabled for the
> package with the dependency, or disabled otherwise.".
>
> The question then is whether that applies only to *DEPEND, or whether
> it's to anywhere where there's a "package with the dependency". Given
> that the section that specifies it isn't limited to *DEPEND, but
> describes dependency specifications in general, I'd say that PMS
> requires it to be supported.
>
> Having said that, it's pretty horrible, and unless there's a good use
> case for such a feature, I'd not be opposed to shoving in a retroactive
> ban on it...
>
tove was trying to use it for perl:
http://git.overlays.gentoo.org/gitweb/?p=proj/perl-overlay.git;a=commitdiff_plain;h=1a4be047e0ad2f13a7eeab9a5295557023ef2d62
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [gentoo-pms] Clarification of has_version cat/pkg[use=] behavior for EAPI 3 or 4
2009-09-12 18:13 ` Petteri Räty
@ 2009-09-12 18:20 ` Ciaran McCreesh
2009-09-12 18:39 ` Petteri Räty
0 siblings, 1 reply; 7+ messages in thread
From: Ciaran McCreesh @ 2009-09-12 18:20 UTC (permalink / raw
To: gentoo-pms
[-- Attachment #1: Type: text/plain, Size: 632 bytes --]
On Sat, 12 Sep 2009 21:13:06 +0300
Petteri Räty <betelgeuse@gentoo.org> wrote:
> tove was trying to use it for perl:
> http://git.overlays.gentoo.org/gitweb/?p=proj/perl-overlay.git;a=commitdiff_plain;h=1a4be047e0ad2f13a7eeab9a5295557023ef2d62
Well, I still think it's horrible (and I think the same about pretty
much anything else using has_version...), but I can't come up with any
creative way of interpreting PMS to say that it *should* be illegal...
So if we're making a change, it's a retroactive one to work around a
Portage bug...
*shrug* I have no strong opinion on how we handle it.
--
Ciaran McCreesh
[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 198 bytes --]
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [gentoo-pms] Clarification of has_version cat/pkg[use=] behavior for EAPI 3 or 4
2009-09-12 18:20 ` Ciaran McCreesh
@ 2009-09-12 18:39 ` Petteri Räty
2009-09-12 19:46 ` Ulrich Mueller
0 siblings, 1 reply; 7+ messages in thread
From: Petteri Räty @ 2009-09-12 18:39 UTC (permalink / raw
To: gentoo-pms
Ciaran McCreesh wrote:
> On Sat, 12 Sep 2009 21:13:06 +0300
> Petteri Räty <betelgeuse@gentoo.org> wrote:
>> tove was trying to use it for perl:
>> http://git.overlays.gentoo.org/gitweb/?p=proj/perl-overlay.git;a=commitdiff_plain;h=1a4be047e0ad2f13a7eeab9a5295557023ef2d62
>
> Well, I still think it's horrible (and I think the same about pretty
> much anything else using has_version...), but I can't come up with any
> creative way of interpreting PMS to say that it *should* be illegal...
> So if we're making a change, it's a retroactive one to work around a
> Portage bug...
>
> *shrug* I have no strong opinion on how we handle it.
>
At least has_version is an improvement over built_with_use.
Going with it being a Portage bug is fine by me.
Regards,
Petteri
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [gentoo-pms] Clarification of has_version cat/pkg[use=] behavior for EAPI 3 or 4
2009-09-12 17:41 [gentoo-pms] Clarification of has_version cat/pkg[use=] behavior for EAPI 3 or 4 Petteri Räty
2009-09-12 18:11 ` Ciaran McCreesh
@ 2009-09-12 19:43 ` Christian Faulhammer
1 sibling, 0 replies; 7+ messages in thread
From: Christian Faulhammer @ 2009-09-12 19:43 UTC (permalink / raw
To: gentoo-pms; +Cc: betelgeuse
[-- Attachment #1: Type: text/plain, Size: 529 bytes --]
Hi,
Petteri Räty <betelgeuse@gentoo.org>:
> I see a couple course of possible actions:
> 1. Current PMS already implies this behavior and Portage is buggy.
> Let's document this explicitly for EAPI 3 and have Portage fixed so
> using EAPI 3 makes sure this syntax works.
Looks like the best practise to me. It is a corner case, but better
be safe.
V-Li
--
Christian Faulhammer, Gentoo Lisp project
<URL:http://www.gentoo.org/proj/en/lisp/>, #gentoo-lisp on FreeNode
<URL:http://gentoo.faulhammer.org/>
[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 198 bytes --]
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [gentoo-pms] Clarification of has_version cat/pkg[use=] behavior for EAPI 3 or 4
2009-09-12 18:39 ` Petteri Räty
@ 2009-09-12 19:46 ` Ulrich Mueller
0 siblings, 0 replies; 7+ messages in thread
From: Ulrich Mueller @ 2009-09-12 19:46 UTC (permalink / raw
To: Petteri Räty; +Cc: gentoo-pms
>>>>> On Sat, 12 Sep 2009, Petteri Räty wrote:
> Ciaran McCreesh wrote:
>> Well, I still think it's horrible (and I think the same about
>> pretty much anything else using has_version...), but I can't come
>> up with any creative way of interpreting PMS to say that it
>> *should* be illegal... So if we're making a change, it's a
>> retroactive one to work around a Portage bug...
> Going with it being a Portage bug is fine by me.
+1
Changing PMS seems unnatural in this case. Why should the USE flag
syntax for has_version be different from the one in dependencies?
Ulrich
^ permalink raw reply [flat|nested] 7+ messages in thread
end of thread, other threads:[~2009-09-12 19:46 UTC | newest]
Thread overview: 7+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2009-09-12 17:41 [gentoo-pms] Clarification of has_version cat/pkg[use=] behavior for EAPI 3 or 4 Petteri Räty
2009-09-12 18:11 ` Ciaran McCreesh
2009-09-12 18:13 ` Petteri Räty
2009-09-12 18:20 ` Ciaran McCreesh
2009-09-12 18:39 ` Petteri Räty
2009-09-12 19:46 ` Ulrich Mueller
2009-09-12 19:43 ` Christian Faulhammer
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox