| 1 |
On Tue, 2009-03-17 at 13:55 +0000, Ciaran McCreesh wrote: |
| 2 |
> On Tue, 17 Mar 2009 10:50:17 +0300 |
| 3 |
> Peter Volkov <pva@g.o> wrote: |
| 4 |
> > If failures are non fatal I don't object to having src_test enabled by |
| 5 |
> > default and I'll all for this even. |
| 6 |
> |
| 7 |
> ...and src_test becomes utterly worthless again. |
| 8 |
|
| 9 |
Your definition of worthless doesn't match the one in my vocabulary, |
| 10 |
because it is very much worth it when many developers have FEATURES=test |
| 11 |
and routinely make sure the packages they use pass the test or file a |
| 12 |
bug. |
| 13 |
|
| 14 |
It is quite irresponsible to enable that by default for the FULL user |
| 15 |
base, given the state of the tree in regards to it, many upstreams |
| 16 |
stance on tests and their failures, and the (very) considerable extra |
| 17 |
time it takes to run them while it's already slower in relation to |
| 18 |
binary distributions. |
| 19 |
|
| 20 |
Enabling tests by default feels like driving users away, because all of |
| 21 |
a sudden their upgrades taken even more time (possibly unexplained to |
| 22 |
them, as an EAPI bump in an ebuild introducing it is not visible to |
| 23 |
them), and they'd just say to hell with it and go to a binary |
| 24 |
distribution that runs the tests for maintainers only, as we should. |
| 25 |
|
| 26 |
Yet we have the _choice_ to take that extra time and double-check on |
| 27 |
maintainers if they really did their job right. |
| 28 |
|
| 29 |
|
| 30 |
-- |
| 31 |
Mart Raudsepp |
| 32 |
Gentoo Developer |
| 33 |
Mail: leio@g.o |
| 34 |
Weblog: http://planet.gentoo.org/developers/leio |
Archives