[forgetfulness]

What really obliterated Scala’s momentum was PySpark overtaking Scala Spark coming from Python’s foothold in Data Science, columnar data warehouses carving out a big chunk of the batch processing pie as well, and then the Akka licensing change.

The Enterprise ecosystem quickly withered away, and now only type level programming diehards remain.

[hocuspocus]

Akka's demise started before the license change. It's an incredible piece of software but as it turns out, not so many people need stateful cluster sharding. Modern cloud architecture and simpler streaming libraries have made the Akka toolkit irrelevant to many use cases.

You're right about Spark and the next logical step will be removing the JVM from the equation entirely, which is already ongoing (see Photon or Comet).

[blandflakes]

I think realistically we're looking at a lot of causes (which is not surprising; it's rarely a singular thing). Scala 3's momentum may have already been negative but how Scala 3 landed represented a nail in the coffin for a lot of individuals, teams, and organizations.

[hocuspocus]

Scala 3 really landed at a bad time as the hype behind Spark, Akka, Play had already weaned off. Typesafe was struggling financially, and didn't back Scala 3 then (that went to VirtusLab).

Typelevel and ZIO really put a lot of effort to release many libraries on day 1 despite the hurdles with the 3.0 release, because these communities are alive and healthy. Yet Odersky is on a small crusade against monadic effect systems, which discouraged a lot of good people. On top of that there are neverending feuds and US politics...