Open staltz opened 1 year ago
Oh interesting. That is good input, only be in the application layer, but still worth considering.
Yeah I'm thinking mostly in SSB URIs
Yes base58 is more or less becoming the standard to encode byte array in URI's
Another benefit of base58 over base64: the padded =
which carries no information ends up wasting precious characters, such that it ends up having the same length (in some cases) as the base58 encoding. Example:
base64:
OAiOTCroL1xFxoCKYaZJDTxhLOHaI1cURm/HSPvEy7s=
base58:
4mjQ5aJu378cEu6TksRG3uXAiKFiwGjYQtWAjfVjDAJW
Same string length.
One of the quirks of working with base64 is that for URIs you have to escape them (
+
=>-
and/
=>_
) and working with this is a pain in the (mini)butt.Other bases like base58 don't have this problem, and the encoding efficiency is almost the same (73% for base58, 75% for base64). Also base58 is used in bitcoin and others, so implementation is not going to be a problem, it's probably provided already for all languages.