3 ms·
> Values are identified by random UUIDs Alternatively and preferably, values are identified by a cryptographic hash.
by jamesfisher 12y ago
> Values are identified by random UUIDs
Alternatively and preferably, values are identified by a cryptographic hash.
- rubyfan 12y agoWhy is that preferable? I'm not even convinced a uuid is appropriate to solve the problem.
- weavejester 12y agoThe proposed protocol allows for PUTs, which means the server needs to atomically check whether or not that UUID is already in the database. Using a cryptographic hash as an ID removes the need for atomicity in the case of PUTs.