| 1 |
On Wed, Jul 27, 2011 at 6:02 AM, Pacho Ramos <pacho@g.o> wrote: |
| 2 |
> El mié, 27-07-2011 a las 09:39 +0200, Michał Górny escribió: |
| 3 |
>> Hello, |
| 4 |
>> |
| 5 |
>> As many of us already raged, the Python eclasses are delaying half |
| 6 |
>> a year with support of EAPI=4. The reason for that is not actually |
| 7 |
>> the lack of time or complexity of needed changes but willingness to use |
| 8 |
>> the new EAPI as an excuse to turn the eclass API upside down. |
| 9 |
>> |
| 10 |
>> The question I'm raising here: should eclasses be actually allowed to |
| 11 |
>> do *heavy* changes in their APIs in such cases? Or should the eclass |
| 12 |
>> maintainers create a new eclass instead (python-r1.eclass or so)? |
| 13 |
>> |
| 14 |
>> The main advantage I see in that is that devs are somehow forced |
| 15 |
>> to migrate their ebuilds as soon as they bump EAPI in them. Taking |
| 16 |
>> a look at a number of ebuilds still using git.eclass (instead of git-2) |
| 17 |
>> this is a serious advantage. |
| 18 |
>> |
| 19 |
>> On the other hand, I find this idea very unclear. Why should two |
| 20 |
>> ebuilds use completely different eclass variables just because they're |
| 21 |
>> using two different EAPIs? More importantly, why is a dev forced to do |
| 22 |
>> the migration in a random point when he/she wants to bump the ebuild |
| 23 |
>> EAPI? I'd like to remind you that python eclass is still hard to read |
| 24 |
>> for many of us. |
| 25 |
>> |
| 26 |
>> And why do we have to wait so long to use a new EAPI? We already had to |
| 27 |
>> fix a lot of ebuilds when old EAPIs were banned in Python eclasses. We |
| 28 |
>> wanted to bump the ebuilds to EAPI 4 then but the eclasses didn't |
| 29 |
>> support it. And now it still doesn't come with EAPI 4 support. |
| 30 |
>> |
| 31 |
>> And keeping two different EAPIs in a single eclass file means probably |
| 32 |
>> that older EAPIs are going to be banned at a random point once again. |
| 33 |
>> Devs will have to pro-actively migrate their ebuilds, overlays will |
| 34 |
>> break and so on. The usual procedure related to eclass removal wouldn't |
| 35 |
>> apply. |
| 36 |
>> |
| 37 |
>> So, don't you think it would be better to simply add EAPI=4 support to |
| 38 |
>> python eclass with no changes and start developing the new API in |
| 39 |
>> python-r1? Devs could migrate then at any point they want (and have |
| 40 |
>> time to!), and when Python team wants to get rid of the old eclass, |
| 41 |
>> the usual removal procedure will apply. |
| 42 |
>> |
| 43 |
> |
| 44 |
> About the concrete case of python eclass, per Arfrever's comment in bug |
| 45 |
> report related with its eapi4 support, that support is already available |
| 46 |
> in overlay, but not yet merged to the tree (probably because of the |
| 47 |
> possible upcoming retirement of Arfrever :-/). What is preventing python |
| 48 |
> team to merge eclass from overlay? |
| 49 |
> |
| 50 |
|
| 51 |
AFAIK, the EAPI4 support in the overlay is EAPI 4-python, that almost |
| 52 |
certainly will never come to gx86, and some guys are trying to port |
| 53 |
the functionality to raw EAPI4, IIRC. |
| 54 |
|
| 55 |
Regards, |
| 56 |
|
| 57 |
-- |
| 58 |
Rafael Goncalves Martins |
| 59 |
Gentoo Linux developer |
| 60 |
http://rafaelmartins.eng.br/ |
Archives