View previous topic :: View next topic |
Author |
Message |
panserg Apprentice
Joined: 16 Apr 2003 Posts: 188
|
Posted: Mon Apr 28, 2003 3:51 am Post subject: Logical expressions in KEYWORDS |
|
|
Let's assume I use ACCEPT_KEYWORDS="ppc" and also that I'd like to update all packages to their max stable version (max version with KEYWORDS="ppc") - no problem:
Now, I want to see what would be updated to unstable (max version with KEYWORDS="~ppc"):
Code: | ACCEPT_KEYWORDS="~ppc" emerge -up world |
Now I want to be somewhere in the middle. I trust the criteria of stability on x86. But I'd like to adjust that stability on ppc to one on x86. In other words I want to participate in testing of "outstanding" package versions, which are already stable on x86, but still unstable on ppc. Basically, I need something like:
Code: | ACCEPT_KEYWORDS="x86 ~ppc" emerge -up world |
In case of gnome a direct "emerge -up ebuild" can be a pain in a neck. Besides, it reminds me RPM, and no way I am gonna back to RPM by myself
The problem is that emerge takes two (or perhaps more) keywords, but it conjuncts them with OR (by picking up the max version for EITHER of them, from UNION of version sets availble for each of keywords), while I want to conjunct them with AND (by picking up the max version for BOTH of them, from the INTERSECTION of version sets availaible for each of keywords).
If there was a reason of using OR-operator in interpreting multiple keywords, than I'd like to have a control of what logical operator to use.
Code: | ACCEPT_KEYWORDS="x86 && ~ppc" emerge -up world |
If I mistake and there is another (elegant!) way to do so then I'd like to know it. If there is a reason of not doing so - stand up and say so. Otherwise, I am thinking to submit an enhancement report in bugzilla.
This post has "ppc" - ONLY AS AN EXAMPLE. It can be sparc, aplha and any other. |
|
Back to top |
|
|
rac Bodhisattva
Joined: 30 May 2002 Posts: 6553 Location: Japanifornia
|
Posted: Mon Apr 28, 2003 7:11 am Post subject: |
|
|
Mixing architectures in ACCEPT_KEYWORDS doesn't make a lot of sense to me. You might end up with binary A-only code being installed on arch B. _________________ For every higher wall, there is a taller ladder |
|
Back to top |
|
|
axxackall l33t
Joined: 06 Nov 2002 Posts: 651 Location: Toronto, Ontario, 3rd Rock From Sun
|
Posted: Mon Apr 28, 2003 11:33 am Post subject: |
|
|
rac wrote: | Mixing architectures in ACCEPT_KEYWORDS doesn't make a lot of sense to me. You might end up with binary A-only code being installed on arch B. |
Binary code for A must have "-" ("minus") prefix for all other architectures (Like "-B"). The proposed use case deals with (A + ~B), where "tilda" means that it should work, but it's not tested enough on B, or it has some temporary minor bugs on B.
It is proposed to make the life of testers better. Besides, the proposed used case of having an explicit logical operator just is developing (actually it's fixing it) what we already have any way - the implicit logical operator OR. Who has decided that OR is better than AND? And Why?
BTW, your comment is inconsistent with what we already have anyway, don't you think so? |
|
Back to top |
|
|
|
|
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum
|
|