| 1 |
On Fri, 2009-05-29 at 10:42 +0000, Duncan wrote: |
| 2 |
> Michael Haubenwallner <haubi@g.o> posted on Fri, 29 |
| 3 |
> May 2009 10:14:46 +0200: |
| 4 |
> |
| 5 |
> > Ohw, the latter would be necessary here, or '4.ebuild' would not be |
| 6 |
> > found. |
| 7 |
> |
| 8 |
> s/4.ebuild/4.eclass/ I assume. |
| 9 |
|
| 10 |
Indeed. |
| 11 |
|
| 12 |
> Except... since an ebuild must presently be sourced to (properly) |
| 13 |
> determine EAPI, it still doesn't work for changes that break sourcing or |
| 14 |
> other pre-EAPI processing (like parsing PN and PVR out of the filename), |
| 15 |
> so the changes allowed with a simple EAPI change are still rather limited. |
| 16 |
|
| 17 |
As long as the *whole* ebuild content (including the filename) needs to |
| 18 |
be successfully interpreted just for EAPI detection, we will have to |
| 19 |
change the extension or wait the extended period for each incompatible |
| 20 |
EAPI change. So this full interpretation for EAPI detection doesn't feel |
| 21 |
like a good way to go at all. |
| 22 |
|
| 23 |
> That's why the change to GLEP55 or alternative, whether in-filename or |
| 24 |
> in-file-itself, will again require either an extended wait after |
| 25 |
> introduction (the old way) or at minimum, a one-time change to extension |
| 26 |
> such that old PM versions don't even see the currently EAPI incompatible |
| 27 |
> changes. |
| 28 |
|
| 29 |
Wouldn't it be possible to avoid both the extension change and another |
| 30 |
extended wait period for new incompatible(*) EAPIs, when we do this |
| 31 |
early and silent exit hack for unsupported ebuilds with old PMs that |
| 32 |
still do full interpretation for EAPI detection? |
| 33 |
|
| 34 |
And after another extended wait period(**), we can start dropping the |
| 35 |
silent exit hack, so we finally have switched to EAPI detection without |
| 36 |
full interpretation, but still have the .ebuild extension. |
| 37 |
|
| 38 |
(*) The incompatibility of EAPIs must not begin (meaning the bytewise |
| 39 |
ebuild content) before the end of both the ebuild's EAPI value |
| 40 |
definition and the silent exit hack. |
| 41 |
But this IMO is an acceptable compromise. |
| 42 |
|
| 43 |
(**) After this wait period, the incompatibility of EAPIs can start |
| 44 |
after the end of the ebuild's EAPI value definition. |
| 45 |
|
| 46 |
Thanks! |
| 47 |
|
| 48 |
/haubi/ |
| 49 |
-- |
| 50 |
Michael Haubenwallner |
| 51 |
Gentoo on a different level |
Archives