Is there a principle issue with concurrency in concatenative languages, or is it simply just missing?
Or am I just missing something?
As far as I know, currently, the most advanced concatenative language is Factor and it still only has a co-operative threading system that doesn't use multiple cores.
Perhaps Slava Pestiv, Factor's inventor, has been consumed by Google so much that it he simply didn't find the time to create a multicore version.
As far as I understand, an actor model should be very fitting for a concatenative language. However, this is a rather difficult area. Any idea what kind of model would work well?
You can think of the 144 cores of the GreenArrays chip as "actors" using message passing between cores. In practice this is exactly how the chip is used; a single problem is broken into small, communicating pieces running concurrently. These are each programmed in Forth which can be thought of as a concatenative language.