| 1 |
On Aug 13, 2017 6:38 AM, "Michael Orlitzky" <mjo@g.o> wrote: |
| 2 |
|
| 3 |
On 08/13/2017 01:01 AM, Hans de Graaff wrote: |
| 4 |
> On Sat, 2017-08-12 at 05:58 -0400, Michael Orlitzky wrote: |
| 5 |
>> |
| 6 |
>> I simply overlooked the global USE change in make.conf because IMO |
| 7 |
>> it's a nonsense operation. |
| 8 |
> |
| 9 |
> This also happens routinely as new python and ruby versions are marked |
| 10 |
> stable, not via make.conf, but by removing their use.stable.mask. |
| 11 |
> |
| 12 |
|
| 13 |
If --newuse is required for that, it shouldn't be. Using a feature of my |
| 14 |
package manager (--newuse) to facilitate a change to a variable on my |
| 15 |
machine (USE) is one thing, but changes to ::gentoo are another. |
| 16 |
|
| 17 |
|
| 18 |
I don't think it is practical or useful to revbump a thousand ebuilds when |
| 19 |
we have a python_targets visibility change, or when the default |
| 20 |
python_targets value changes. |
| 21 |
|
| 22 |
While revbumping might be a good idea for specific IUSE changes, I don't |
| 23 |
think it is always necessary. |
Archives