The article mentions KSUID which has a very similar spec, but with a wider random segment.
For comparison: KSUID is 160-bit wide versus ULID sticks to 128-bit wide (which is somewhat more compatible with UUID-like database column types, for instance, so long as they don't try things like UUID version checks). KSUID uses Base62 encoding versus ULID uses a much simpler to (lexicographicly) sort Base32 encoding.
(I've found ULID useful in some of my own projects.)
Comments
I have stopped using UUID and GUID in favor of https://github.com/ulid/spec
The article mentions KSUID which has a very similar spec, but with a wider random segment.
For comparison: KSUID is 160-bit wide versus ULID sticks to 128-bit wide (which is somewhat more compatible with UUID-like database column types, for instance, so long as they don't try things like UUID version checks). KSUID uses Base62 encoding versus ULID uses a much simpler to (lexicographicly) sort Base32 encoding.
(I've found ULID useful in some of my own projects.)