12:03 <discocaml> <myrlhex> Is there a way to refer to the parent with `(include_subdirs qualified)`

12:03 <discocaml> <myrlhex> I'm making a compiler, and I wanted to put all the passes in one directory.

12:04 <discocaml> <myrlhex> Oh wait, it works now

12:04 <discocaml> <myrlhex> You can just use the full module name (the `X__Y` variant)

14:53 <discocaml> <fleabagwithouthumor> I'm new to OCaml and I find the suffix notation for constructors interesting, is there a reason behind this choice ? (e.g. `'a option` compared to `Maybe a` in Haskell)

14:57 <discocaml> <lukstafi> `'a option` is French for `Maybe a`

15:00 <discocaml> <fleabagwithouthumor> Is the reason behind it only because it's inria and they're French ?

15:00 <discocaml> <fleabagwithouthumor> I mean it's possible

15:03 <discocaml> <lukstafi> C also uses the backward syntax, e.g. `int *`.

15:05 <discocaml> <lukstafi> https://en.wikipedia.org/wiki/ML_(programming_language)#List_reverse shows the same syntax as OCaml

15:05 <discocaml> <lukstafi> So that falsifies my hypothesis...

15:07 <discocaml> <hockletock> I believe ML also uses places type variables in front

15:09 <discocaml> <lukstafi> Yes https://www.pure.ed.ac.uk/ws/portalfiles/portal/15143545/1_s2.0_0022000078900144_main.pdf

15:10 <discocaml> <lukstafi> I think the motivation comes from analogy with the function type.

15:12 <discocaml> <fleabagwithouthumor> as in `'a option` is `'a -> option` ?

15:12 <discocaml> <fleabagwithouthumor> I think that's actually a good way to read it, thanks

15:13 <discocaml> <limp.biskit> maybe also to differentiate between variants?

15:16 <discocaml> <fleabagwithouthumor> I have no idea how variants are defined, I think I've come far enough with what I know about haskell, time to RTFM !

15:16 <discocaml> <fleabagwithouthumor> I have no idea how variants are written, I think I've come far enough with what I know about haskell, time to RTFM !

15:52 <discocaml> <deepspacejohn> I think it's also supposed to be similar to in English you would say "int list" to mean a list of ints

15:54 <Armael> yea, in fact in French "list int" is more natural than "int list"

15:54 <Armael> so I'd blame the ML tradition here

15:56 <discocaml> <deepspacejohn> https://discuss.ocaml.org/t/job-build-system-engineer-at-jane-street/8737/7?u=johnj

16:19 <discocaml> <limp.biskit> variants are written like `Some_variant <val>`, so i imagine if you could also have `some_type <typ>` it may be a bit less clear than it is now?

16:19 <discocaml> <limp.biskit> just a guess

16:23 <discocaml> <smondet> yes the ML tradition comes from english "conteinee container"

16:23 <discocaml> <smondet> 'int list' 'cigar box' 'shit hole'

16:25 <discocaml> <smondet> it's actually later, that haskell turned (some) containers into adverbs: Maybe, Either, ...

17:04 <dh`> ocaml has four different argument syntaxes and there's no need for any of that, it's just far too late to change

17:06 <discocaml> <deepspacejohn> I vaguely remember seeing someone say that they teach students to just always use the GADT syntax for regular ADTs because apparently `:` is easier for them to remember than `of`? or something like that.

17:32 <discocaml> <contextfreebeer> is there a way to pass data to the top level parsing function in Menhir? I know you can functorize the grammar but I'm not sure if you can pass first class modules to a functor, and that sounds pretty hacky anyway as I only want to pass a record with some options from the command line

17:37 <discocaml> <contextfreebeer> if not it can be worked around in any case

17:51 <discocaml> <deepspacejohn> Not that I'm aware of. It may make sense to do another pass to apply options after parsing.

17:52 <discocaml> <contextfreebeer> yeah, that is what I'm going to do, I think that's a better approach anyway

17:54 <discocaml> <deepspacejohn> yeah, I think that's probably the better practice.

17:58 <dh`> you can make the start token return a function

17:58 <dh`> but that's probably gross

18:35 <discocaml> <stephanosf.> Has anyone tried running ocaml on virtualbox ? Had some issues with ocaml on windows (dont know how to get latest version after default DkML installation) and was thinking on switching to either dualboot or a virtual machine

18:39 <discocaml> <leviroth> I use ocaml on windows via WSL2, which I think is a fairly popular approach.

18:39 <discocaml> <27theo> why not WSL (window subsystem for linux)?

18:42 <discocaml> <stephanosf.> I also read through wsl2 but i was still confused on if you are able to use versions after 4.14.0 , i am trying to utilize the shared memory capabilities of the 5.0.0 release

18:42 <discocaml> <stephanosf.> https://cdn.discordapp.com/attachments/436568060288172044/1233488357099573412/image.png?ex=662d470d&is=662bf58d&hm=0c6b157d225f861b9343024ea14a48b38c5a8443eccb049c064e801bc3f612c5&

18:43 <discocaml> <leviroth> Where does that table come from?

18:43 <discocaml> <stephanosf.> https://ocaml.org/docs/ocaml-on-windows

18:44 <discocaml> <stephanosf.> does wsl2 support later versions after 4.14.0 for you @therewolf ?

18:44 <discocaml> <leviroth> Yes

18:44 <discocaml> <leviroth> Tbh I don't understand what the stuff about tiers is

18:45 <discocaml> <stephanosf.> damn cause ive been using dkml and i couldnt get to versions after 4.14.0 , do you use vscode to right ocaml code though? ( i reade somewhere that its not yet compataible with the latest ocaml versions)

18:45 <discocaml> <stephanosf.> write*

18:52 <discocaml> <regularspatula> the vscode works with latest ocaml versions

18:54 <discocaml> <stephanosf.> Alright ill give it a try , thanks guys

22:11 <discocaml> <limp.biskit> using 5.1 on wsl2 fine

22:12 <discocaml> <limp.biskit> wsl2 is just a linux vm, it supports any version you can run on linux

22:24 <discocaml> <stephanosf.> Alright ill give it a try thank you @limp.biskit

22:29 <discocaml> <limp.biskit> will have to compile it with opam though, don’t know of any distros shipping ocaml 5 binaries