Could just be the status page software itself. It looks like it uses https://github.com/louislam/redbean-node which is kind of cursed

> This automatically generates the tables and columns... on-the-fly. It infers relations based on naming conventions.

I haven’t really found that free services scale the same way. It’s hard for the “open source community” to contribute and improve the quality of bottlenecks that are only encountered by one operator.

When you take OSS projects that scale well, say Linux, Postgres, Kafka, redis, etc. they either scale up (Linux) which is arguable easier, or were able to scale out because there are thousands, if not millions, that have massive deployments pushing them to their limits.

Unless there is some sort of secure way to “open source” operational data for codeberg, or many others running huge deployments of Forgejo I don’t see it being very effective.

I do see Google having another go at code hosting though.