My intention was not to check the actual SHA1 in anything other
than t0000-basic.sh, but some test in t4000 series were buggy
and ended up comparing the SHA1.
I have since corrected them and I believe the test suite in the
current tip of Linus tree has the corrected compare_diff_raw and
friends in t/diff-lib.sh, which t4000 series use to avoid this
problem. I just hand munged COPYING and README and ran t4008 in
the Linus tree, and do not see the problem you are reporting.
Could you verify that your test failure is coming from the
version that was before t/diff-lib.sh was introduced? Otherwise
I have to investigate this further.
Hi, Junio!
On Tue, 2005-05-31 at 21:27 -0700, Junio C Hamano wrote:
My intention was not to check the actual SHA1 in anything other
than t0000-basic.sh, but some test in t4000 series were buggy
and ended up comparing the SHA1.
I have since corrected them and I believe the test suite in the
current tip of Linus tree has the corrected compare_diff_raw and
friends in t/diff-lib.sh, which t4000 series use to avoid this
problem. I just hand munged COPYING and README and ran t4008 in
the Linus tree, and do not see the problem you are reporting.
Could you verify that your test failure is coming from the
version that was before t/diff-lib.sh was introduced? Otherwise
I have to investigate this further.
The failure in on the "cogito" branch
(rsync://rsync.kernel.org/pub/scm/cogito/cogito.git), which doesn't have
t/diff-lib.sh.
The "git" branch (rsync://www.kernel.org/pub/scm/git/git.git) is fine.
Sorry for the noise. I should have checked that branch too.
--
Regards,
Pavel Roskin