Riven v3 tracking issue

[MingweiSamuel]

• Generate schema code in Rust, possibility for (semi-)automatic updating refactor: generate schema code in build.rs #83
  • Do we want to keep the generated code tracked in the main branch, or if not how can we have nice diffs between generations? (generated in a different branch?)
• Figure out an API to unify the new try_ methods from feat: add try_ methods requiring min_capacity to be available to execute ("minimum quota feature") #79
• Cleanup the error types (maybe use thiserror as a helper)
  • Remove source_reqwest_error
• Adjust rate limiting
  • Move away from using sleep "rejection sampling" to obtain rate limits, use a proper waker queue/semaphore instead
  • Examine possible race conditions when obtaining rate limit tokens due to not holding on to locks
• Figure out what other breaking changes to make!

[BMorinBlitz]

Adjust rate limiting

Why was sliding log chosen over token bucket? (see Governor https://github.com/boinkor-net/governor/blob/master/governor/src/state/in_memory.rs)