feat: close event error code
- Closes #59
@emilk i do not think adding reason string is good now. unfortunately there is some variability in codes and specs and likely impls, making adding it brittle over several platforms.
i think can cleanup comments and code just for code, but for reason imho need to think more (by thinking i mean tokio and web sys teams think about things as experts - web-sys::CloseEvent, tokio::CloseFrame and RFC - all have different inconsistent interfaces - i am afraid cannot rely on same behavior of reason on all stacks)
reason doesn't need to be consistent, just be human-readable.
Btw, merge in main in order to (hopefully) fix the CI
@emilk i can finish PR only with code, no string. if need string - please close PR.
i do no see that string abstraction is consistent in web sys and tungestine. also i biased to custom error codes (so that each category+source of error - own code - so no need strings).