That’s really great to hear, that things will be moving along in such short order. Do you have any good introductions to the topic of algebraic effects, beyond what I could get by searching on my own?
Although GIL seems like a huge limitation, I’ve been actually quite impressed by how infime it was for most tasks (except if you do require huge throughput).
Folks, I recall reading somewhere that OCaml has something like a GIL and threaded/paralleled apps are a bit of an issue. Is this still true? Thanks!
Yes there is a GIL and parallelism is currently doable with c bindigs as in python, but there is ongoing work for multicore (and algebraic effects) support: https://discuss.ocaml.org/t/ocaml-multicore-report-on-a-june-2018-development-meeting-in-paris/2202
The work is still ongoing, and went through some iterations in the past few years, but the first bits should start landing by the end of the year
That’s really great to hear, that things will be moving along in such short order. Do you have any good introductions to the topic of algebraic effects, beyond what I could get by searching on my own?
I would recommend http://okmij.org/ftp/Haskell/extensible/index.html
And for more specifically ocaml:
Although Ocaml ignored multicore a long time, the Standard ML community keeps doing interesting projects in that space. MultiMLton is one example.
Although GIL seems like a huge limitation, I’ve been actually quite impressed by how infime it was for most tasks (except if you do require huge throughput).