The number one problem with commercial OCaml is the lack of comprehensive library support. Until that's addressed, it's not going to take off. Brilliant language, not very practical. F# is a step in the right direction here.
I wonder if there's any hope of ever getting type classes in OCaml. Then you'd be able to have stuff like a single "print" function you can give any type of value that can be turned into a string representation.
It would be interesting to know what they plan to do. It's easy to speculate l. It would be much more interesting to see ocaml on the jvm
A decade ago it looked like O'Caml was going to be the language that brought FP to the masses. It had good runtime performance, good tools, and a decent library. Then the development team basically did no work on the language for some eight or so years, and other languages, notably Scala, F#, and Haskell, have gained ground.
It is hard to cross from academia to commerce. The changes that were (and are) needed to O'Caml are not things that will result in publications, and thus don't fit the academic funding model. At the same time they are a precursor to attracting wider interest in the language. It seems that a benefactor like Jane Street is needed for this to occur. Certainly it seems that Jane Street is eager to spend money to fix this problem.
The story of O'Caml is a good reminder that in language adoption, as in business, execution, along with a bit of luck, what matters. The world is full of could-have-beens. And on that note, I've got code to write...