| 1 |
> If, as a user or an arch person, I get a src_test failure right now, I |
| 2 |
> don't know whether this means "eek! Something's gone wrong, and I |
| 3 |
> really need to fix this" or "oh, whoever maintains this package |
| 4 |
> doesn't care". But with EAPI 2, I'll be able to know that a src_test |
| 5 |
> failure really does mean "something's wrong". |
| 6 |
> |
| 7 |
|
| 8 |
A developer should always care about src_test in my opinion. That some |
| 9 |
developers don't is a nuisance, and the core of the problem. But trying |
| 10 |
to change this via EAPI is doing things the wrong way round: At first |
| 11 |
there has to be an agreement about the importance of test suites (backed |
| 12 |
up by strict policies that every developer is bound to), and then we can |
| 13 |
discuss if it makes sense to activate them by default in EAPI-X. |
Archives