| 1 |
On Sun, Dec 6, 2015 at 2:03 PM, Michał Górny <mgorny@g.o> wrote: |
| 2 |
> Hi, |
| 3 |
> |
| 4 |
> Here's another patchset. Mostly fixups inspired by multilib-build.eclass |
| 5 |
> changes with a few little additions. |
| 6 |
> |
| 7 |
> Changes: |
| 8 |
> |
| 9 |
> 1. eclass-set variables are now read-only, |
| 10 |
|
| 11 |
Sounds ok, but might break some ebuilds. |
| 12 |
|
| 13 |
> 2. 'unset -f' is used to unset temporary & local functions, |
| 14 |
> 3. implementations are reordered for sane order. |
| 15 |
> |
| 16 |
> I've tested this with a few dozen random distutils-r1, python-r1, |
| 17 |
> python-any-r1 and python-single-r1 packages. However, for |
| 18 |
> the implementation reorder a larger tinderbox run would be appreciated. |
| 19 |
> |
| 20 |
> As explained in the commit, the reorder may influence files installed by |
| 21 |
> a package, and implementation selected by python_setup(). This should |
| 22 |
> not cause issues for correctly written ebuilds, and should help us find |
| 23 |
> those that are not correctly written ;-). |
| 24 |
> |
| 25 |
> In other words, we're finally considering Python 3.x preferred over |
| 26 |
> Python 2.x. |
| 27 |
|
| 28 |
That could be a significant change, and I do expect that this will |
| 29 |
break some ebuilds. |
| 30 |
|
| 31 |
Any easy way to test this first? Or should we just be ready for the bug reports? |
Archives