2007/9/11, Junio C Hamano <gitster@pobox.com>:
> I also suspect this would not pass on CRLF boxes.

I finally removed the hardcoded object IDs (patch attached,
to be applied on top of the previous patch).

But I don't know what to fix for making the test to pass
in CRLF boxes. Could it due to the comparison of
expected diffs with files created using <<EOF?