tags:

views:

203

answers:

1

Some time ago I came across a website that described a binary-to-text encoding which was human friendly. For example on input (for decoding) it accepted 0, o and O all as the same value, because people tend to mix those characters very easy.

Unfortunately I can't remember how it was called...

Edit: this was the one I was looking for: "Another alternative design for Base32 is created by Douglas Crockford, who proposes using additional characters for a checksum."

+4  A: 

Sounds like Base32.

There are variants (such as z-base-32) which accept O for 0, l/I for 1, etc.

Colin Pickard
The zero '0' character is not part of the Base32 character set.
anon
Reading the link it seems that the poster wants one of the variations of Base32 rather than the RFC standard version.
Dave Webb
yeah, in the wiki link it mentions z-base-32 which covers 0/O 1/l/I
Colin Pickard
Neither Base32 nor z-base-32 solve all the problems: there is still potential to mix-up '5' and 'S'.
Dan