26

Re: Why did not fly up OCaml?

There is a judgement that many lispo-functional languages are created first of all because of the academic interest. 1) we take / we give birth to a formalism 2) we research regularities 3) we write articles / to the dissertation 4)... 5) PROFIT!!!

27

Re: Why did not fly up OCaml?

Hello, monax, you wrote: M>F# is, consider, OCaml. Only they added there null and connected with.net. Like as people get accustomed and even use, but for me there is a big lack - MS. I.e. under  I in any way will not get it ( it is not counted, this strange thing) And what in Mono strange? By the way, F# it is tested including on Mono on Linux. And the considerable part of users use it on such platform.

28

Re: Why did not fly up OCaml?

Hello, nikov, you wrote: N> And what in Mono the strange? By the way, F# it is tested including on Mono on Linux. And the considerable part of users use it on such platform. And generally, at what here  when.NET core more year works on .

29

Re: Why did not fly up OCaml?

Hello, monax, you wrote: M> And here it became interesting, and why it did not fly up and did not become one of mainstream languages? As well as in evolutionary selection, the niche where it could develop is necessary. Read absence of alternative.

30

Re: Why did not fly up OCaml?

Hello, Kodt, you wrote: Frenchmen - perverts, at first  SML Well, not that that they "" SML. OCaml there is from a dialect ML th appeared to SML and syntactically is more similar on first ML than SML which is join Kordellevsky VAX ML and Hope. Then started to push in it any interesting features, and with absolutely eclectic syntax, and without thinking about an infrastructure and community. Here only SML in which features too tried to push through, but they rested in Milnera and stuck - much more .

31

Re: Why did not fly up OCaml?

Hello, monax, you wrote: M> But at all these fine characteristics in a mainstream it is not present, well unless in a type F#, but it more recently. So why it did not win the place in the sun? Generally, at  many serious lacks, but at mainstream languages much or is still worse. For example, at  toy  without SMP, but at popular dynamic languages , as a rule, even more poor and toy. I think that it is not necessary to search for the reason (not) popularity in any properties of language. Any evidences of that they to see somehow connected to me it to was not necessary.

32

Re: Why did not fly up OCaml?

Hello, monax, you wrote: M> Haskel another absolutely. How much I understood, lazy calculations in  puzzle to people because to force it to create side-effect difficult. However, it while only on responses, it was not possible to study it yet. Ghost effects in  form easily. Simply they should be registered explicitly through types, and it is unusual for many. M>F# is, consider, OCaml. Only they added there null and connected with.net. Like as people get accustomed and even use, but for me there is a big lack - MS. I.e. under  I in any way will not get it ( it is not counted, this strange thing), and under  I just and work. So for passage on F# there should be something not ordinary that I spat all knowledge in a linuh-ecosystem and passed in . Here or me it will be necessary to carry out the task which here 100 % well lay down on +, but does not lay down at all on +, or F# should offer to me such wild advantages that without it in any way. But it will not be, and  is "with a view of erudition increase", on  it did not gather to drag yet. F# as language is pleasant to me even more than Ocaml. Especially in the first are pleasant sequence expressions and its generalization computations expressions. But too the knot on fat   afflicts.

33

Re: Why did not fly up OCaml?

Hello, Klapaucius, you wrote: K> Generally, at  many serious lacks, but at mainstream languages much or is still worse. For example, at  toy  without SMP, but at popular dynamic languages , as a rule, even more poor and toy. To  I did not reach yet. And what with it not so?

34

Re: Why did not fly up OCaml?

Hello, monax, you wrote: M> To  I did not reach yet. And what with it not so? The main lack does not support SMP, simultaneously can will be fulfilled only one flow. That is all that is connected to flows works (the same Thread unit) but as as if it is fulfilled on the single-processor and one-nuclear computer. The same lack is we tell at a python from it GIL, or at nodejs. An output to use instead of flows processes. The second big lack limitation of many types for example max_int for 32 bits only 1073741823, one bit of the data is taken away  by system for GC. Even more strongly restriction in sizes for arrays Sys.max_array_length only 4194303. All it certainly already is a little significant for 64 bits, but for 32 it is a noticeable lack. The third lack which strongly affected popularity on mine, it  on unix figurative systems, windows is faster by a residual principle though recently with it it is better. Well and plus very poor standard library. It is a little corrected by library http://batteries.forge.ocamlcore.org/