Base58 is a encoding that has an alphabet with 58 letters. As 58 is not a base of 2 (like 32 or 64) this
encoding is based on the div-mod algorithm (big integer arithmetic). Therefore encoding and decoding is not as
efficient as with e.g.
Base64. Therefore, it should not be used for very large binary data but is fine for
hashes, IDs, etc.
Base58 is typically a good compromise of usability and storage-efficiency:
- like
Base32 it typically uses alphabets that only contain Latin letters and digits and can also avoid
similar characters (like '1' and 'l' as well as 'O' and '0'). Therefore encoded representations can be selected with
a simple double-click and if you have to manually read and type them there are no ambiguous or very similar looking
characters.
- its overhead in length of the encoded representation compared to the original binary data is better than
Base32 (that also has the above benefits) but only a little worse than Base64.