[gentoo-dev] Use of use.mask

[gentoo-dev] Use of use.mask

[Ole Markus With]

[-- Attachment #1: Type: text/plain, Size: 911 bytes --]

Hi all,

I was thinking of adding SVN snapshot ebuilds of PHP to the tree.
Ebuilds for PHP extensions use USE_EXPAND to decide which slots (and
thus, which ABIs) of PHP the extension should be built for, much like
ruby does. A new USE_EXPAND USE flag should therefore be added for the
SVN snapshot slot. The problem is that the ABI is not stable and should
only be used by people who 'know what they are doing', and the snapshot
ebuilds will probably always be without keywords. This will cause some
dependency problems.

The only solution I could think of would be to add this new USE flag to
use.mask. But as far as I could tell, use.mask is meant for masking USE
flags that do not work on certain architectures etc. It is also a bit
tricky for users to unmask USE flags.

Is this still the best way to do this? Or are there any better ways that
I did not think of?

Cheers,
Ole Markus


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 554 bytes --]

Re: [gentoo-dev] Use of use.mask

[Anthony G. Basile]

On 04/30/2011 02:47 PM, Ole Markus With wrote:
> Hi all,
> 
> I was thinking of adding SVN snapshot ebuilds of PHP to the tree.
> Ebuilds for PHP extensions use USE_EXPAND to decide which slots (and
> thus, which ABIs) of PHP the extension should be built for, much like
> ruby does. A new USE_EXPAND USE flag should therefore be added for the
> SVN snapshot slot. The problem is that the ABI is not stable and should
> only be used by people who 'know what they are doing', and the snapshot
> ebuilds will probably always be without keywords. This will cause some
> dependency problems.
> 
> The only solution I could think of would be to add this new USE flag to
> use.mask. But as far as I could tell, use.mask is meant for masking USE
> flags that do not work on certain architectures etc. It is also a bit
> tricky for users to unmask USE flags.
> 
> Is this still the best way to do this? Or are there any better ways that
> I did not think of?
> 
> Cheers,
> Ole Markus
> 

I don't see that this is much different in philosophy than p.masking
experimental/broken ebuilds which we add to the tree for dev only
testing.  In both cases a user who thinks they 'know what they're doing'
can locally unmask, at their own risk.

-- 
Anthony G. Basile, Ph.D.
Gentoo Linux Developer [Hardened]
E-Mail    : blueness@gentoo.org
GnuPG FP  : 8040 5A4D 8709 21B1 1A88  33CE 979C AF40 D045 5535
GnuPG ID  : D0455535