| 1 |
>>>>> On Sat, 7 Nov 2015, David Leverton wrote: |
| 2 |
|
| 3 |
> Well, a big use case stated in bug 282296 is package.mask in |
| 4 |
> overlays, which I assume refers to the repo-level one rather than |
| 5 |
> one inside an actual profile. If people haven't made it clear that |
| 6 |
> they want that, I think it's because of imprecise language rather |
| 7 |
> than because they don't actually want it. |
| 8 |
|
| 9 |
> (And this has been argued over for long enough - I'd really rather |
| 10 |
> not have it go on ever further when people realise the spec still |
| 11 |
> doesn't allow what they're trying to do.) |
| 12 |
|
| 13 |
This is all well and good, but the patch with (essentially) the |
| 14 |
current wording is attached to bug 282296 since 18 months and it talks |
| 15 |
only about files in a profile, but not about the profiles/ directory. |
| 16 |
I think nobody can claim that these 18 months have been too short a |
| 17 |
time for a careful review. |
| 18 |
|
| 19 |
Therefore, I see only the alternatives that either the feature will be |
| 20 |
accepted with the current wording, or that the feature will be dropped |
| 21 |
from EAPI 6 altogether. |
| 22 |
|
| 23 |
What is *not* going to happen is that we delay EAPI 6 because of this. |
| 24 |
Also I am strongly opposed against any substantial changes on the day |
| 25 |
before the council meeting. There was plenty of time to get things |
| 26 |
right, and if there are still such major issues at the current point, |
| 27 |
then regrettably this will have to wait for EAPI 7. |
| 28 |
|
| 29 |
Ulrich |
Archives