-
-
-
InfluxDB
Power Real-Time Data Analytics at Scale. Get real-time insights from all types of time series data with InfluxDB. Ingest, query, and analyze billions of data points in real-time with unbounded cardinality.
still works. In my experience so far, you only need to worry if you define your own combinators, e.g. like in https://github.com/phadej/github/commit/dd3dfbccaf4de4c3095b52fefb52c6b3c961a8f8
Ah, I see you are working on this already, thank you: https://github.com/haskell/aeson/pull/883
I benchmarked my jose library (using the benchmark suite from https://github.com/marcin-rzeznicki/libjwt-typed, which uses criterion); the JSON objects involved are small (< 8 members) and the performance difference is negligible - perhaps slightly faster (don't have to allocate a vector of hash buckets, most of which are unused). I haven't benchmarked performance with huge numbers of members but it's O(1) [amortised, degrading to O(n) for pathological inputs] -> O(log n), so I would expect a small performance decrease for objects with >> 8 members.