| 1 |
Hi, |
| 2 |
|
| 3 |
13.09.14 22:03, Peter Stuge написав(ла): |
| 4 |
>> E.g. we in downstream have some patches, when upstream changes |
| 5 |
>> related code (e.g. applying our patches), ebuild fails to build. |
| 6 |
> |
| 7 |
> I consider this a separate issue however. I would actually expect |
| 8 |
> there to be a policy which forbids patches on live ebuilds. Make |
| 9 |
> another live ebuild or maybe an overlay if you want to offer a |
| 10 |
> different set of commits than the upstream repo. |
| 11 |
> |
| 12 |
|
| 13 |
In the ideal country of elves. In the real life it can be not possible to build and install software in a given distribution without downstream patches. You can find examples of such live ebuilds in Gentoo tree. |
| 14 |
|
| 15 |
> |
| 16 |
>> Everything in live ebuild can change, so it will fail. |
| 17 |
>> It can be not crap, but some improvements, but it does not matter |
| 18 |
>> for the possibility of building of a given ebuild. |
| 19 |
> |
| 20 |
> Patches aside I completely agree that a live ebuild is a moving |
| 21 |
> target, but to me that is a good thing, in fact the very essence |
| 22 |
> of open source. |
| 23 |
> |
| 24 |
|
| 25 |
Open source is not about instability. It is just about sources being available. No more, no less. ;-) |
| 26 |
|
| 27 |
Anyway, summarizing, it is completely impossible to be sure that live ebuild will be buildable for you on a given arch in the next 15 min., even if it was so in the last 15 min. |
| 28 |
|
| 29 |
-- |
| 30 |
Jauhien |
Archives