Adoption of the OSS version must not have been very high, otherwise I would have expected a Valkey / OpenTofu style, community-led fork.

I'm guessing battle-tested reliability isn't a priority for calendaring/scheduling web services, unlike Redis/Valkey.

It's probably cleaner for anyone looking to adapt the source code to point an LLM at it to extract some specs and tests, then build a new one from scratch.