[merge][rfc] don't re-run tests in LANG=C from 'make check'?

[John Arbash Meinel]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Martin Pool wrote:
| === modified file 'Makefile'
| --- Makefile    2007-12-17 01:45:32 +0000
| +++ Makefile    2008-02-07 05:29:54 +0000
| @@ -32,8 +32,6 @@
|
|  check: docs extensions
|         $(PYTHON) -Werror -O ./bzr selftest -1v $(tests)
| -       @echo "Running all tests with no locale."
| -       LC_CTYPE= LANG=C LC_ALL= ./bzr selftest -1v $(tests) 2>&1 |
| sed -e 's/^/[ascii] /'
|
|
| The second test roughly doubles the pqm merge time because of running
| the test suite twice.  (Not quite double, because the compilation and
| the merge itself is done only once.)
|
| I think this is unnecessary and should be removed: the majority of
| tests are not particularly dependent on the locale.  When there is
| something unicode-dependent, we should be trying to test it in a way
| that does not depend on the tester's locale - I would say if this ever
| fails, we should add a new test that is specific to that bug.
|
| In a way we might as well have all the coverage we can get, but the
| pqm run time does slow things down, particularly at a release.
|
| I don't recall ever seeing a merge that passed the first stage and
| failed at the second.  But other people merge more often than I have
| recently, and if you have found it useful I will leave it in.
|

The only reason I can think of are for Unicode tests, where the person
forgets that they need to allow for platforms that don't support it, and
need to be skipped there.

John
=:->
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (Cygwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFHqp8DJdeBCYSNAAMRAqCeAJwJ7q6d3C1RAjjhqHJuvmKwjdom1QCgkjhE
zHXt55Kg/CB7499giHjAkXg=
=bgxN
-----END PGP SIGNATURE-----