Follow-up Comment #2, bug #61594 (project make):
It's a bit confusing because the "good and useful points" were all made in
responses to the mailing list, not posted to this Savannah issue.
In any event, I don't see any point in using any sort of cryptographic hash,
even one as widely available as SHA-1. All that's wanted here, from my
understanding, is a value which is highly likely to be unique. It seems that
something like xxhash or the GNU hash-pjw function would suffice. These,
however, give results that are only 4 or 8 bytes wide as opposed to something
like 16 bytes for MD5 or 20 bytes for SHA-1.
I don't know if these extra bytes are worth the cost.
In any event any such function would need to allow a way to choose the type of
hash to avoid lock-in: we wouldn't want to just have "$(hash <somestring>)"
then never be able to change it, I wouldn't expect.
_______________________________________________________
Reply to this item at:
<https://savannah.gnu.org/bugs/?61594>
_______________________________________________
Message sent via Savannah
https://savannah.gnu.org/
