nicoo has quit [Remote host closed the connection]
nicoo has joined #ocaml
jrslepak has quit [Ping timeout: 256 seconds]
ShalokShalom has quit [Ping timeout: 240 seconds]
jbrown has quit [Ping timeout: 264 seconds]
pierpal has joined #ocaml
ShalokShalom has joined #ocaml
pierpal has quit [Read error: Connection reset by peer]
pierpal has joined #ocaml
jnavila has quit [Ping timeout: 240 seconds]
pierpal has quit [Ping timeout: 240 seconds]
ShalokShalom has quit [Ping timeout: 263 seconds]
shinnya has joined #ocaml
muelleme has joined #ocaml
muelleme has quit [Ping timeout: 240 seconds]
ShalokShalom has joined #ocaml
muelleme has joined #ocaml
muelleme has quit [Ping timeout: 264 seconds]
cbot has joined #ocaml
jack5638 has quit [Ping timeout: 240 seconds]
jack5638 has joined #ocaml
cbot_ has quit [Ping timeout: 240 seconds]
hdon has joined #ocaml
cbot has quit [Quit: Leaving]
hdon has quit [Ping timeout: 260 seconds]
mathZ has joined #ocaml
mathZ has quit [Remote host closed the connection]
pierpal has joined #ocaml
steenuil has joined #ocaml
TEPEMTAK has joined #ocaml
* TEPEMTAK
ping sagax
<sagax>
ага, привет
jimt has quit [Quit: jimt]
pierpal has quit [Ping timeout: 256 seconds]
mfp has joined #ocaml
exarkun has quit [Read error: Connection reset by peer]
ia0 has quit [Quit: reboot]
TEPEMTAK has quit []
exarkun has joined #ocaml
TEPEMTAK has joined #ocaml
* TEPEMTAK
ping sagax
ia0 has joined #ocaml
Gurkenglas has joined #ocaml
<Leonidas>
octachron: there is a `make_fake_id()` in your code, where I would add a space for clarity
* TEPEMTAK
попалкал тыкочкой sagax
<sagax>
>_>
<Leonidas>
octachron: And I don't know if talking about "weak typing" is a good idea, I find it possibly more confusing than helpful than to write out the entire name.
<sagax>
i did not found something like as "ocaml for 15 minutes", maybe anybody know where i can find this
<Leonidas>
octachron: yes, to make it look less like Reason syntax
<Leonidas>
octachron: you mind if I post it to lobste.rs?
<sagax>
hm, what different between list and array? in short
<moolc>
попалкал... тыкочкой... ухты...
<Leonidas>
sagax: a list is a linked list of element with Cons of 'a | Nil, whereas an array is a contingous region of memory, like an array in C.
<Leonidas>
Could we please dial down the russian? It is hard to engage in all kinds of discussions when they span over french, russian and english :)
<sagax>
thanks, of course i will reading about this in book or docs
<octachron>
Leonidas, maybe a good compromise would be "weak type variable" rather than "weak type", "weakly polymorphic type variable" is a bit of a mouthful
<octachron>
Leonidas, would you mind posting the OCaml manual link? Or at least, give me the time of updating the disclaimer.
<Leonidas>
octachron: Yes, that's better. My concern was mainly that people after skimming over it might walk away with the impression OCaml supports weak typing (in the C sense), which is about the opposite.
<Leonidas>
octachron: Sure, I can wait until it is incorporated in the manual :)
<pmetzger>
I hate the green in the current manual design but perhaps for the moment match it
<pmetzger>
?
<pmetzger>
maybe slightly darker because it will be isolated rather than used as a background.
<pmetzger>
So far as I can tell, the distinction between "design" and "production" in those URLs is ~160 bytes vs. ~210 bytes.
<pmetzger>
"production" has whitespace etc removed.
<steenuil>
I think just making them green like that wouldn't make them stand out from the background
<steenuil>
and the ones that are there right now are yellow with a black border
<steenuil>
I guess I'll have to try later
<pmetzger>
Yah, you're right. But plain yellow would disappear against the white
<pmetzger>
Red would also be garish.
<pmetzger>
I'm not sure.
<pmetzger>
Should I leave this to you for today?
sh0t has joined #ocaml
<steenuil>
I'm kind of busy with some other stuff but maybe I'll try later
gareppa has joined #ocaml
webshinra_ is now known as webshinra
mk9 has joined #ocaml
cYmen has joined #ocaml
<cYmen>
Can somebody recommend a tutorial, please? The website has this ominous warning about everything being out of date. :p
al-damiri has joined #ocaml
<pmetzger>
Yes. dev.realworldocaml.org is a great intro book and is online.
<cYmen>
pmetzger: thank you! :)
<exarkun>
Does opam determine what modules a library exports? jbuilder? Or only the mli files?
<pmetzger>
look at a typical opam file. :) it more or less only describes how something is built and what it depends on, but not any detail like that. the .mli files that get installed are what really determines things.
pierpal has quit [Read error: Connection reset by peer]
<exarkun>
That's basically the impression I was getting... So maybe I should ask my real question instead, sorry. I have a project with a bunch of source directories and a bunch of mli files. Mostly, I can't get utop to see any of the things in the mli files.
<exarkun>
I installed stuff with opam (from a git checkout) and I can `#require ...` the libraries. But to no apparent effect.
<exarkun>
I don't understand how ... stuff ... get resolved.
sgnb has joined #ocaml
<exarkun>
(nor what to read to learn how it works)
pierpal has joined #ocaml
<pmetzger>
Did you reload the environment after doing the opam magic?
<exarkun>
I started a new shell and eval'd `opam config env`
<pmetzger>
That's all you should need.
<pmetzger>
I've now exhausted my meagre supply of ideas.
<exarkun>
`opam list` shows the libraries I think should be exporting modules (as "dev (pinned)")
<Armael>
exarkun: did #require ... succeed, but then you are unable to use the module? or did it fail?
<exarkun>
#require succeeded. utop even offered the library names for autocomplete.
<exarkun>
After that, I expected that I'd be able to use types and values in the mli files in the source for the library.
<exarkun>
But they're all unbound
<pmetzger>
did you open the modules?
<exarkun>
no
<pmetzger>
did you try using the names of the modules when mentioning a module component, "Foo.bar"?
<Armael>
um, eg utop advertises "Foo.bar", but if you type "Foo.bar;;" it says unbound?
<exarkun>
Armael: no, in the #require command it offered "tezos-p2p"
pierpal has quit [Quit: Poof]
<exarkun>
and then I'm not really sure what should happen next. "tezos-p2p" is not a valid identifier, of course.
pierpal has joined #ocaml
<exarkun>
so what module would its identifiers be in?
<Armael>
ah
<Armael>
well in theory it can be anything
<Armael>
in practice it should be something like Tezos_p2p
<pmetzger>
Look at the .mli files. :)
<def`>
ls `ocamlfind query tezos-p2p`
<exarkun>
pmetzger: What do I look in the mli files for?
<exarkun>
def`: ah, interesting.
<pmetzger>
exarkun: Are you quite new to OCaml? If so: .mli files define what symbols a module exports.
<exarkun>
pmetzger: But I think I have some grasp of mli files.
<def`>
the type information are stored in cmi files
<Armael>
mli are not necessarily installed
<def`>
so: ocamlfind query tezos-p2p gives you the directory where the library is installed
<pmetzger>
.cmi is after compilation, yah. Sorry about not explaining that right. If there's no other provision made, the name of the file the .mli/.cmi file has is used to derive the module name.
<def`>
mli files are human readable and not used by utop, but you can read to understand what the library offers
<exarkun>
That gist is the result of "ls `ocamlfind query tezos-p2p`" in case that wasn't clear
<def`>
cmi files are used by utop, give the contents of this directory you should be able to access Tezos_p2p (tezos_p2p.cmi)
<exarkun>
what the...
<exarkun>
Well, I *definitely* tried `#require "tezos-p2p"` without useful results before.
<pmetzger>
So for example, tezos_p2p__P2p.cmi implies a module named Tezos_p2p__P2p under the rules, I think.
<exarkun>
But _this_ time, utop happily offers all kinds of Tezos_... for autocomplete.
<def`>
pmetzger: the "__" modules are aliases that are not meant to be accessed directly
<exarkun>
I have no idea what I did differently this time.
<def`>
tezos_p2p__P2p can be accessed through Tezos_p2p.P2p in practice
<pmetzger>
def`: Gotcha. I find a lot of that stuff opaque still.
<def`>
pmetzger: I agree :P
mk9 has quit [Read error: Connection reset by peer]
<exarkun>
Okay, so this helps a lot. Thanks guys.
<pmetzger>
def`: When I ask the compiler to create a first pass .mli file for me it often spits those names out and I have to clean it up based on what I know the names are supposed to be. This is irritating at times.
<exarkun>
/gals/whatever.
<pmetzger>
(I've only been using OCaml for six months or so.)
<cYmen>
"unbound module Base", apt says ocaml-base is installed. what is my mistake?
<pmetzger>
cYmen: if you're using dev.realworldocaml.org, I would suggest you install opam with apt, and then use opam to install a recent ocaml and the tools. the site has instructions, I'll link to them in a moment.
<cYmen>
I'm still not okay with the fact that every programming language comes with a package manager these days. :)
<pmetzger>
It really only takes a couple of minutes btw.
<exarkun>
cYmen: +1
<pmetzger>
Yah, well, it would be nice if we didn't need one for every language but we don't live in that world. :(
gareppa has quit [Quit: Leaving]
<pmetzger>
cYmen: what OS are you on?
<cYmen>
currently ubuntu
<pmetzger>
So you're in good shape. You install opam, opam init, install a bunch of packages very quickly, you should be up and running in five minutes.
<pmetzger>
What's your preferred editor?
<cYmen>
yeah, it's not a problem
<pmetzger>
vim or emacs?
<cYmen>
I'm fine with both.
<pmetzger>
I'm not a vim guy, so no good advice there, but in Emacs you'll want tuareg+merlin, it really makes things nicer. There are instructions on that page.
<pmetzger>
With a tiny amount of configuration, you'll know when you have syntax errors as you're typing, will know when there are type errors immediately, etc. It's kind of magical.
<pmetzger>
By the time you actually want to run code you know it will actually run. Sadly it won't debug for you.
hdon has joined #ocaml
<cYmen>
hehe
gareppa has joined #ocaml
<Armael>
there's a package for automatic editor setup
<cYmen>
Armael: got a link?
<Armael>
it's the opam package "user-setup"
<Armael>
you can try to install it after you install opam
FreeBirdLjj has quit [Remote host closed the connection]
FreeBirdLjj has joined #ocaml
nullifidian has quit [Remote host closed the connection]
jnavila has joined #ocaml
<exarkun>
Hm. merlin-locate stopped working.
cbot has joined #ocaml
<exarkun>
What do I have to do beyond launch my emacs into an opam-enabled environment to let merlin work right?
<pmetzger>
Emacs configuration is a sufficiently turing equivalent thing that it's hard to say.
rain1 has left #ocaml ["Leaving"]
<moolc>
nice understatement
FreeBirdLjj has quit [Remote host closed the connection]
FreeBirdLjj has joined #ocaml
muelleme has joined #ocaml
FreeBirdLjj has quit [Ping timeout: 264 seconds]
nullifidian has joined #ocaml
pierpal has quit [Ping timeout: 240 seconds]
exarkun has quit [Read error: Connection reset by peer]
pierpal has joined #ocaml
Guest66217 is now known as Haudegen
nullifidian has quit [Remote host closed the connection]
exarkun has joined #ocaml
kakadu has quit [Quit: Konversation terminated!]
nullifidian has joined #ocaml
_etc has joined #ocaml
<_etc>
Hi! I'm working with the compiler frontend and the Path and Longident modules both seem pretty similar — with variants ident,dot,apply. Does anyone know the logic of when a path is used and when an lident is used? Also is there a fn to convert between the two? Thanks!
<_etc>
Maybe it's that lidents are for parsetrees and paths are used internally for typing?
hdon has joined #ocaml
<pmetzger>
_etc: I suspect that asking on discuss.ocaml.org rather than here might catch more of the right people.
<_etc>
pmetzger: thanks! I'll try that
pierpa has joined #ocaml
<def`>
_etc: Longident are just syntax
<def`>
(it is just a structured way to represent "A.B.c")
<def`>
Path have more semantic information: it is a longident that has been resolved by the compiler.
<def`>
The root of a Path is an identifier, rather than a string, which is name augmented with some unique information relative to the context in which it is being used.
<def`>
which is a name augmented...*
jnavila has quit [Ping timeout: 256 seconds]
eh_eff has joined #ocaml
<_etc>
def`: Ah that makes sense — thanks!
malina has quit [Remote host closed the connection]
shinnya has joined #ocaml
_etc has quit [Ping timeout: 260 seconds]
Denommus has joined #ocaml
_etc has joined #ocaml
muelleme has quit [Ping timeout: 240 seconds]
nullifidian has quit [Remote host closed the connection]
tarptaeya_ has joined #ocaml
tarptaeya has quit [Ping timeout: 240 seconds]
malina has joined #ocaml
<_etc>
Does anyone know if there's a way to find all the free-variables in a parsetree?
<_etc>
or a typed tree would work!
shinnya has quit [Ping timeout: 240 seconds]
eh_eff has quit [Ping timeout: 248 seconds]
kakadu has joined #ocaml
jnavila has joined #ocaml
jbrown has quit [Ping timeout: 265 seconds]
nullifidian has joined #ocaml
<octachron>
_etc, if you include open, you probably need to pull a good portion of the typechecker
Haudegen has quit [Read error: Connection reset by peer]
moolc has left #ocaml ["ERC (IRC client for Emacs 27.0.50)"]
<exarkun>
"Tezos_p2p.P2p.create" on the rhs of = is what's underlined by utop, fwiw
pierpal has quit [Ping timeout: 240 seconds]
<octachron>
exarkun, labels (the name of the named arguments) are part of the type
<exarkun>
in the source the labels aren't a, b, c, or d, they're "meta" and "msg" but matching those names doesn't change the error (except for changing the names in it)
nicoo has quit [Remote host closed the connection]
nicoo has joined #ocaml
Haudegen has joined #ocaml
<exarkun>
oh
<exarkun>
sorry
<exarkun>
I'll try to read more carefully in the future.
<exarkun>
Thank you
<octachron>
no problem, the labels are easy to miss when you don't expect them to pop up in type error messages
_etc has quit [Ping timeout: 260 seconds]
gtrak has quit [Ping timeout: 268 seconds]
eh_eff has joined #ocaml
muelleme has joined #ocaml
muelleme has quit [Ping timeout: 264 seconds]
cbot has quit [Quit: Leaving]
gtrak has joined #ocaml
gtrak has quit [Ping timeout: 260 seconds]
gtrak has joined #ocaml
jnavila has quit [Remote host closed the connection]
steenuil has quit [Remote host closed the connection]
mk9 has joined #ocaml
_andre has quit [Quit: leaving]
argent_smith has quit [Quit: Leaving.]
Bronsa` has joined #ocaml
mk9 has quit [Quit: mk9]
mk9 has joined #ocaml
tarptaeya_ has quit [Quit: Konversation terminated!]
pmetzger has quit []
Gurkenglas has quit [Read error: Connection reset by peer]
gtrak has quit [Ping timeout: 248 seconds]
gtrak has joined #ocaml
spew has quit [Quit: Connection closed for inactivity]
kakadu has quit [Remote host closed the connection]
gtrak has quit [Ping timeout: 248 seconds]
gtrak has joined #ocaml
nomicflux has joined #ocaml
gtrak has quit [Ping timeout: 240 seconds]
caltelt has joined #ocaml
Soni has quit [Excess Flood]
Soni has joined #ocaml
malina has quit [Remote host closed the connection]
eh_eff has quit [Quit: WeeChat 1.7]
malina has joined #ocaml
sz0 has joined #ocaml
theglass has quit [Ping timeout: 256 seconds]
malina has quit [Quit: Throwing apples of Montserrat]
mathZ has joined #ocaml
mathZ has quit [Remote host closed the connection]
mathZ has joined #ocaml
sh0t has quit [Remote host closed the connection]
malina has joined #ocaml
mathZ has quit [Remote host closed the connection]