Closed bwesterb closed 1 year ago
Hashing is very fast already
In other words, never use MTCs with SHA-3? 😉
Probably the right padding construction would be
struct {
opaque issuer_id<1..32>;
uint32 batch_number;
opaque pad[N];
uint8 distinguisher = ...;
...
That way you can feed the first block into your hash context once and then clone the hash context for all the subsequent bits. Previously we needed to freeze three different ones.
In other words, never use MTCs with SHA-3? 😉
shush!
Hashing is very fast already: we do not need the speed up, and the padding made it a bit harder to implement.
GH #72