| 1 |
On Sat, 5 Aug 2006 11:49:53 +0200 |
| 2 |
Danny van Dyk <kugelfang@g.o> wrote: |
| 3 |
|
| 4 |
> Please re-read the list of packages that fail tests: |
| 5 |
> * glibc |
| 6 |
> * autoconf |
| 7 |
> * gettext |
| 8 |
> * tar |
| 9 |
> That makes _4_ system packages. Before I would consider making |
| 10 |
> FEATURES=test a default, I would add least want the system set to |
| 11 |
> actually merge with it. |
| 12 |
|
| 13 |
So you're happy to let users install these packages without them |
| 14 |
knowing the tests would fail? |
| 15 |
|
| 16 |
I certainly agree they should pass their tests. autoconf-2.60, |
| 17 |
gettext-0.15 and tar-1.15.1-r1, which are the latest versions I |
| 18 |
have installed here, all pass on my system. If they fail on your |
| 19 |
platform, then you should make sure bugs are open and the relevant |
| 20 |
maintainers are doing something about it, and IMO they should not go to |
| 21 |
arch (i.e. should remain ~arch) until the test issues are resolved. |
| 22 |
|
| 23 |
Thing is, at the moment you have a bunch of packages installed that |
| 24 |
fail their tests. This may mean the tests are broken, however it may |
| 25 |
also mean the packages are not working correctly on your system, and |
| 26 |
I'd be concerned if I were you. Avoiding the test phase doesn't make |
| 27 |
the packages work, obviously. |
| 28 |
|
| 29 |
glibc is somewhat of a special case; it is especially sensitive to |
| 30 |
the environment - many of the tests assume a vanilla RedHat |
| 31 |
environment, and often the test failures in glibc are not actual |
| 32 |
problems with glibc but limitations of the test suite. However we |
| 33 |
should not be encouraging people to install glibc versions where the |
| 34 |
test failures are not understood. Clearly if something in glibc is not |
| 35 |
behaving properly, the effects can be nasty. |
| 36 |
|
| 37 |
-- |
| 38 |
Kevin F. Quinn |
Archives