A module for quickly generating IDs guaranteed to be unique within the current process, for the lifetime of the current process.

1. This module should be highly performant even on highly concurrent systems.
2. While this module can run out of unique IDs, this is very unlikely assuming a sane threading model and will panic rather than potentially reusing unique IDs.

API Docs: https://stebalien.github.io/snowflake/snowflake/

Limits

The unique ID's are sizeof(usize) + 64 bits wide and are generated by combining a usize global counter value with a 64bit thread local offset. This is important because each thread that calls new() at least once will reserve at least 2^64 IDs. So, the only way to run out of IDs in a reasonable amount of time is to run a 32bit system, spawn 2^32 threads, and claim one ID on each thread. You might be able to do this on a 64bit system but it would take a while... TL; DR: Don't create unique IDs from over 4 billion different threads on a 32bit system.