| 1 |
Am Donnerstag, 29. März 2018, 15:34:05 CEST schrieb Michał Górny: |
| 2 |
> |
| 3 |
> 1. Does deprecation really mean anything in terms of profiles? Even |
| 4 |
> in the context of EAPI bans we explicitly stated that it affects new |
| 5 |
> packages and EAPI bumps. I think deprecating it for ebuilds is still |
| 6 |
> meaningful even if profiles would stay EAPI 5. |
| 7 |
|
| 8 |
If we state that we deprecate *ebuilds* then it doesnt affect profiles. |
| 9 |
So it just depends on the wording. |
| 10 |
|
| 11 |
Having an EAPI deprecated for part of its impact is kinda messy though. |
| 12 |
|
| 13 |
> |
| 14 |
> 2. Do we want to keep profiles EAPI 5 indefinitely? If we consider it |
| 15 |
> a goal to reduce the number of EAPIs in use, I think it would be |
| 16 |
> reasonable to bump profiles to EAPI 6 proactively, even if it doesn't |
| 17 |
> change anything. |
| 18 |
|
| 19 |
We can bump them, but very slowly... and best at release-like steps like 13.0 |
| 20 |
-> 17.0. |
| 21 |
|
| 22 |
If your portage doesn't understand some ebuilds, that's kinda half-safe for a |
| 23 |
transition time (it just won't see the updates). |
| 24 |
If your portage doesn't understand your profile, that makes things explode |
| 25 |
much more efficiently. |
| 26 |
|
| 27 |
-- |
| 28 |
Andreas K. Hüttel |
| 29 |
dilfridge@g.o |
| 30 |
Gentoo Linux developer |
| 31 |
(council, toolchain, perl, libreoffice, comrel) |
Archives