ULIDs solve these issues. Basically 48 bits of millisecond timestamp then 80 bits of randomness. The technique has been used in MSSQL apps since about 2003 (a

ULIDs solve these issues. Basically 48 bits of millisecond timestamp then 80 bits of randomness. The technique has been used in MSSQL apps since about 2003 (although people called them COMBs instead of ULIDs back then) https://github.com/oklog/ulid https://github.com/oklog/ulid