The short SHA-1 collision test employs two "magic" values in commit
messages to trigger an "ambiguous SHA-1 error" during 'rebase -i' -- one
for 'collide3' during setup, and one for 'collide2' at rebase time --
even though the collision can just as easily be triggered by a single
magic value at rebase time only. The setup-time 'collide3' magic value
serves no purpose [1], but can easily mislead readers into thinking that
it is somehow significant even though it is not. Drop this unneeded bit
of magic.

As a side-effect, this also simplifies the setup-time "test_commit
collide3" invocation, making it easier to read.

[1]: The magic value on 'collide3' gave it a short SHA-1 prefix of
"badbeef" which could be spotted easily in verbose output during
development of the test, but is otherwise not meaningful or helpful.

Signed-off-by: Eric Sunshine <>
 t/ | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/t/ b/t/
index 7d15c7a..50e22b1 100755
--- a/t/
+++ b/t/
@@ -1058,7 +1058,7 @@ test_expect_success 'short SHA-1 setup' '
        unset test_tick &&
        test_commit collide1 collide &&
        test_commit --notick collide2 collide &&
-       test_commit --notick "collide3 115158b5" collide collide3 collide3
+       test_commit --notick collide3 collide
@@ -1069,7 +1069,7 @@ test_expect_success 'short SHA-1 collide' '
        unset test_tick &&
        test_tick &&
        set_fake_editor &&
-       FAKE_COMMIT_MESSAGE="collide2 815200e" \
+       FAKE_COMMIT_MESSAGE="collide2 ac4f2ee" \
        FAKE_LINES="reword 1 2" git rebase -i HEAD~2

To unsubscribe from this list: send the line "unsubscribe git" in
the body of a message to
More majordomo info at

Reply via email to