On Mon, Sep 19, 2022 at 09:07:27AM -0700, Junio C Hamano wrote:
Junio C Hamano [off-list ref] writes:
quoted
... But use of _by_name() that does
not limit the input to hexadecimal _is_ a problem ...
Ah, no, sorry, this was wrong. The original used "rev-parse -q --verify"
without restricting the "single word" to "sha1 prefix".
Yeah, I've sent a v2 that removes the false verdicts from the commit message.
Even if we restricted it to SHAs it would still be ambiguous.. I guess it
doesn't matter in practice.
Thanks