[gentoo-dev] USE_EXPAND additions

[gentoo-dev] USE_EXPAND additions

[Jason Stubbs]

Add FEATURES, ARCH, USERLAND to USE_EXPAND
http://bugs.gentoo.org/show_bug.cgi?id=82513

FRITZCAPI_CARDS in USE_EXPAND is needed for newer fritzcapi versions
http://bugs.gentoo.org/show_bug.cgi?id=84422


There's bound to be more of these come up, but unlikely to be many more 
versions in the portage 2.0.51 series. Thinking about it, these don't really 
belong in portage's domain anyway. The latter should be freely available for 
anybody to do. The former really needs discussion between QA and the arch 
teams (and portage by extension of that discussion - fex multilib).

Anyway, any objections against moving the current USE_EXPAND out of 
make.globals and into base's make.defaults? Those using <=2.0.50* won't get 
any additions (how it is now anyway) and anybody using a stacked profile 
(which requires >=2.0.51) will get whatever is in make.globals overwritten 
with whatever is in make.defaults.

Regards,
Jason Stubbs
--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] USE_EXPAND additions

[Donnie Berkholz]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Jason Stubbs wrote:
> Anyway, any objections against moving the current USE_EXPAND out of 
> make.globals and into base's make.defaults? Those using <=2.0.50* won't get 
> any additions (how it is now anyway) and anybody using a stacked profile 
> (which requires >=2.0.51) will get whatever is in make.globals overwritten 
> with whatever is in make.defaults.

Sounds like a great idea. Power to the people!

Thanks,
Donnie
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (GNU/Linux)

iD8DBQFCXV/fXVaO67S1rtsRAn15AKDcwhXWjd6bGjpNkJET0bHcYtTywwCg4ERP
sQeXjUwSBThlheu+O7Zxi1g=
=XO/F
-----END PGP SIGNATURE-----
--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] USE_EXPAND additions

[Kito]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1


On Apr 13, 2005, at 1:57 PM, Stephen Bennett wrote:

> On Wed, 2005-04-13 at 20:48 +0900, Jason Stubbs wrote:
>> Anyway, any objections against moving the current USE_EXPAND out of
>> make.globals and into base's make.defaults? Those using <=2.0.50* 
>> won't get
>> any additions (how it is now anyway) and anybody using a stacked 
>> profile
>> (which requires >=2.0.51) will get whatever is in make.globals 
>> overwritten
>> with whatever is in make.defaults.
>
> Sounds good to me. Waiting on new portage releases for stuff we want to
> use in ebuilds kinda sucks, so (up to a point) the more we can move 
> into
> profiles the better, as far as I'm concerned.

Agreed.

>
> --
> gentoo-dev@gentoo.org mailing list
>
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (Darwin)

iD8DBQFCXWtmJ0rMK/3OwgsRAjOCAJ0UpDEVomjnctPEa5mtza+MYUZi3ACfagTV
YxaEZaaVpB7TUz3O1IjeLv8=
=BePN
-----END PGP SIGNATURE-----

--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] USE_EXPAND additions

[Stephen Bennett]

On Wed, 2005-04-13 at 20:48 +0900, Jason Stubbs wrote:
> Anyway, any objections against moving the current USE_EXPAND out of 
> make.globals and into base's make.defaults? Those using <=2.0.50* won't get 
> any additions (how it is now anyway) and anybody using a stacked profile 
> (which requires >=2.0.51) will get whatever is in make.globals overwritten 
> with whatever is in make.defaults.

Sounds good to me. Waiting on new portage releases for stuff we want to
use in ebuilds kinda sucks, so (up to a point) the more we can move into
profiles the better, as far as I'm concerned.

--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] USE_EXPAND additions

[Jason Stubbs]

On Wednesday 13 April 2005 20:48, Jason Stubbs wrote:
> Anyway, any objections against moving the current USE_EXPAND out of
> make.globals and into base's make.defaults? Those using <=2.0.50* won't get
> any additions (how it is now anyway) and anybody using a stacked profile
> (which requires >=2.0.51) will get whatever is in make.globals overwritten
> with whatever is in make.defaults.

Just to be specific on this, the corner case is users still using a flat 
profile that upgrade to 2.0.51.20. Those users won't have USE_EXPAND active 
_at all_. The flat profile cannot be fixed for this because of a bug in 
2.0.50 that'll make it traceback on any operation.

Regards,
Jason Stubbs
--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] USE_EXPAND additions

[Jason Stubbs]

On Thursday 14 April 2005 23:26, Jason Stubbs wrote:
> On Wednesday 13 April 2005 20:48, Jason Stubbs wrote:
> > Anyway, any objections against moving the current USE_EXPAND out of
> > make.globals and into base's make.defaults? Those using <=2.0.50* won't
> > get any additions (how it is now anyway) and anybody using a stacked
> > profile (which requires >=2.0.51) will get whatever is in make.globals
> > overwritten with whatever is in make.defaults.
>
> Just to be specific on this, the corner case is users still using a flat
> profile that upgrade to 2.0.51.20. Those users won't have USE_EXPAND active
> _at all_. The flat profile cannot be fixed for this because of a bug in
> 2.0.50 that'll make it traceback on any operation.

If this actually scared anybody (that received it), my apologies. I was 
confused about bugs. Fixing the flat profiles will be problem either. The bug 
is that 2.0.50 will die if two profiles in a cascade define the same 
variable.

Regards,
Jason Stubbs
--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] USE_EXPAND additions

[Ciaran McCreesh]

[-- Attachment #1: Type: text/plain, Size: 606 bytes --]

On Fri, 15 Apr 2005 22:14:59 +0900 Jason Stubbs <jstubbs@gentoo.org>
wrote:
| If this actually scared anybody (that received it), my apologies. I
| was  confused about bugs. Fixing the flat profiles will be problem
| either. The bug  is that 2.0.50 will die if two profiles in a cascade
| define the same  variable.

Which should be fine, because no-one is even going to *think* about
wanting sub-profile-specific USE_EXPAND. Right?

-- 
Ciaran McCreesh : Gentoo Developer (Vim, Fluxbox, shell tools)
Mail            : ciaranm at gentoo.org
Web             : http://dev.gentoo.org/~ciaranm


[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] CONFIG_PROTECT* and USE_EXPAND additions

[Jason Stubbs]

On Friday 15 April 2005 22:18, Ciaran McCreesh wrote:
> On Fri, 15 Apr 2005 22:14:59 +0900 Jason Stubbs <jstubbs@gentoo.org>
>
> wrote:
> | If this actually scared anybody (that received it), my apologies. I
> | was  confused about bugs. Fixing the flat profiles will be problem
                                                         ^not^a^
> | either. The bug  is that 2.0.50 will die if two profiles in a cascade
> | define the same  variable.
>
> Which should be fine, because no-one is even going to *think* about
> wanting sub-profile-specific USE_EXPAND. Right?

USE_EXPAND is not an incremental.

The only other settings that directly relate to ebuilds and currently live in 
make.globals are CONFIG_PROTECT and CONFIG_PROTECT_MASK. These should also be 
moved into the tree. Any objections?

Regards,
Jason Stubbs
--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] CONFIG_PROTECT* and USE_EXPAND additions

[Ciaran McCreesh]

[-- Attachment #1: Type: text/plain, Size: 715 bytes --]

On Fri, 15 Apr 2005 23:20:43 +0900 Jason Stubbs <jstubbs@gentoo.org>
wrote:
| USE_EXPAND is not an incremental.

Indeed (despite what Nick has said to the contrary). Nor should it be --
down that path lies insanity, and not the good kind.

| The only other settings that directly relate to ebuilds and currently
| live in  make.globals are CONFIG_PROTECT and CONFIG_PROTECT_MASK.
| These should also be  moved into the tree. Any objections?

Only so long as you stick a big fat "# DO NOT MODIFY THESE WITHOUT
DISCUSSION ON -dev" on them.

-- 
Ciaran McCreesh : Gentoo Developer (Vim, Fluxbox, shell tools)
Mail            : ciaranm at gentoo.org
Web             : http://dev.gentoo.org/~ciaranm


[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] USE_EXPAND additions

[Paul de Vrieze]

[-- Attachment #1: Type: text/plain, Size: 557 bytes --]

On Friday 15 April 2005 15:14, Jason Stubbs wrote:
> If this actually scared anybody (that received it), my apologies. I was
> confused about bugs. Fixing the flat profiles will be problem either. The
> bug is that 2.0.50 will die if two profiles in a cascade define the same
> variable.

Then make sure that users that have this situation can not install the new 
portage. pkg_setup is not beautifull, but in this case it's probably best.

Paul

-- 
Paul de Vrieze
Gentoo Developer
Mail: pauldv@gentoo.org
Homepage: http://www.devrieze.net

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] USE_EXPAND additions

[Brian Harring]

On Fri, Apr 15, 2005 at 08:53:41PM +0200, Paul de Vrieze wrote:
> On Friday 15 April 2005 15:14, Jason Stubbs wrote:
> > If this actually scared anybody (that received it), my apologies. I was
> > confused about bugs. Fixing the flat profiles will be problem either. The
> > bug is that 2.0.50 will die if two profiles in a cascade define the same
> > variable.
> 
> Then make sure that users that have this situation can not install the new 
> portage. pkg_setup is not beautifull, but in this case it's probably best.
Or just depend on the appropriate portage version instead of sticking 
a die in pkg_setup...
~brian
--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] USE_EXPAND additions

[Jason Stubbs]

On Saturday 16 April 2005 03:53, Paul de Vrieze wrote:
> On Friday 15 April 2005 15:14, Jason Stubbs wrote:
> > If this actually scared anybody (that received it), my apologies. I was
> > confused about bugs. Fixing the flat profiles will be problem either. The
> > bug is that 2.0.50 will die if two profiles in a cascade define the same
> > variable.
>
> Then make sure that users that have this situation can not install the new
> portage. pkg_setup is not beautifull, but in this case it's probably best.

I'm not sure I get you. There's no requirement for using a new portage that I 
can see as the flat profiles can have USE_ORDER added to them as well.

Regards,
Jason Stubbs
--
gentoo-dev@gentoo.org mailing list