tests pass???????????

o dear it seems like it works

but it works

just through logs... turns out it was actually working before all this test stuff? ok neat.

...by bailing

nice

scanning counts-only will be a bit more expensive, but we *can* at least avoid deserializing the HLL if needed, since DbConcat has the (cheap) count as a prefix.

anyway, fmt

hourly truncated cursor

make it make sense

simplified events are in one file now

un-fun time refactoring jetstream to use serde_json RawValue, it's gross and has options everywhere now but it works. bleh.

....

and sketch out record trimming

Ufos deadlock debug

seems to work!

now it's just OOMs and L0 compaction stalls to figure out...

but still locks up

what!

on read. need to actually write rollups still.

fix retry delay reporting, increase max retries

entirely remove the ping loop, it was not doing much

- bug from jetstream oxide: https://github.com/videah/jetstream-oxide/pull/9
- locally just merging create and update into a single variant, switching on `operation` in client
- mod cursor seeks with bounds instead of prefix.next, since post prefixes will be delete tombstones
- the exclusive bound is *super* annoying. this pulls lsm-tree in but that was not even nice. should ask fjall to expose the upper bound calculation as a distinct function?