Allow failed tests to be ignored using make's "-i". The patch also
disables parallel make in t/. This doesn't make the testing any
different as before: the tests were run sequentially before.
It also allows to run more tests, ignoring the ones usually failing
just to figure out if something else broke. (Or to ignore plainly
uninteresting situations because of the testing being done on say...
cygwin ;)
---
How about hiding "ok ..." output?
Alex Riesen [off-list ref] writes:
... (Or to ignore plainly
uninteresting situations because of the testing being done on say...
cygwin ;)
Could you tell us which ones break on Cygwin?
Junio C Hamano, Tue, Nov 08, 2005 20:29:59 +0100:
quoted
... (Or to ignore plainly uninteresting situations because of the
testing being done on say... cygwin ;)
Could you tell us which ones break on Cygwin?
of course. I remember t4000-diff-format having problems, using chmod +x
and comparing the results afterwards (it's not really cygwin problem,
but FAT). I let the whole suite running tomorrow and just post the
failed here (instead of dreaming how nice would it be to live in a
perfect world).
On 11/8/05, Alex Riesen [off-list ref] wrote:
Junio C Hamano, Tue, Nov 08, 2005 20:29:59 +0100:
quoted
quoted
... (Or to ignore plainly uninteresting situations because of the
testing being done on say... cygwin ;)
Could you tell us which ones break on Cygwin?
So far:
t4000-diff-format.sh
* FAIL 3: validate git-diff-files -p output.
compare_diff_patch current expected
t4006-diff-mode.sh
* FAIL 3: verify
diff -u expected check
t4102-apply-rename.sh
* FAIL 3: validate
test -f bar && ls -l bar | grep "^-..x......"