Rainbench is designed to test the throughput of different reactive libraries. It's a simple benchmark that fills a bucket with raindrops.

Each rain drop creates a subcription and the observable gets updated every millisecond, which should trigger a rebuild of each rain drop widget with its new position.

The number of raindrops and the capacity of the bucket can be adjusted. The benchmark will run until the bucket is full. The time it takes to fill the bucket is recorded and displayed at the end of the benchmark.

Results

• Release mode on linux
• Benchmark is restarted after each run to prevent GC interference
• value_LB == value_listenable_builder
• stream_builder and value_LB are not perfect comparisons as they can only listen to 1 observable at a time... but they serve as a good baseline

20k raindrops and a bucket capacity of 30k

Library	Raindrops/s	Time to fill bucket
state_beacon	6112	4.91s
value_LB	5301	5.66s
mobx	4463	6.72s
signals	1868	16.06s
stream_builder	1597	18.78s
context_watch	1245	24.01s
solidart	DNF	DNF

Video:

20knew.mp4