csakatok_ has quit [Remote host closed the connection]
derek_c has joined #ocaml
mcclurmc has joined #ocaml
derek_c has quit [Quit: leaving]
derek_c has joined #ocaml
skchrko has joined #ocaml
hkBst has quit [Ping timeout: 260 seconds]
hkBst has joined #ocaml
hkBst has quit [Changing host]
hkBst has joined #ocaml
mfp has joined #ocaml
ttamttam has quit [Quit: ttamttam]
void64 has joined #ocaml
yacks has joined #ocaml
Simn has joined #ocaml
mcclurmc has quit [Quit: Leaving.]
AltGr has joined #ocaml
<AltGr>
Hi there
<AltGr>
is there a reason why `let module M = (val m : S with type 'a elt = a) in` doesn't parse ?
<AltGr>
it seems you can't specify `with type` constraints with polymorphic types in first-class modules
madroach has joined #ocaml
<AltGr>
there is probably a good reason for that, but I wasn't expecting a syntax error ?
quelu has quit [Ping timeout: 246 seconds]
sgnb has quit [Read error: Connection reset by peer]
sgnb has joined #ocaml
chambart has joined #ocaml
<Anarchos>
AltGr let module M = (module m : S ....) in ...
sgnb has quit [Read error: Connection reset by peer]
sgnb has joined #ocaml
<AltGr>
Anarchos: it parses if I remove the type constraint, or even if I remove the `'a` parameter
<AltGr>
(but I always get the module/val wrong for first-class modules also ;))
quelu has joined #ocaml
<Anarchos>
AltGr did you look the official syntax chapter in the manual ?
<Anarchos>
it often gives clues on the syntax error
<Anarchos>
where does dypgen install the dyplib ?
<Anarchos>
when i run it , it complains about the unknown module Dyp which is installed on <standard dir of ocaml lib>/dyp/
<adrien_oww>
ocamlfind list
<adrien_oww>
does it show anything related to dypgen?
<Anarchos>
no
<Anarchos>
maybe i should install dypgen in teh standard lib/ocaml instead in lib/ocaml/dyp ?
<adrien_oww>
NO!
<adrien_oww>
it seems that unfortunately dypgen doesn't install through ocamlfind
<adrien_oww>
considering where it's installed, you can add "-I +dyp" to the command-line
<Anarchos>
ok this one works : dypgen --pv-obj --ocamlc "-I +dyp" prop.dyp
<Anarchos>
but what a pain to add "-I +dyp" when it is in standard location !!
<adrien_oww>
it's not
<adrien_oww>
and I wouldn't call that a "pain"
<Anarchos>
adrien_oww i should isntall all the .cma files into /lib/ocaml ?
<adrien_oww>
7 chars
<adrien_oww>
NO
<adrien_oww>
that's a good way to fuck up an install, write crap in build systems and make it unusable by others
<Anarchos>
adrien_oww yes i know but i thought dypgen was able to find itself without help :)
ttamttam has joined #ocaml
<adrien_oww>
I prefer that it doesn't
<adrien_oww>
that would hinder ocamlfind support (which it should really be doing =/ )
ttamttam has quit [Client Quit]
chambart has quit [Ping timeout: 276 seconds]
<Anarchos>
everybody here praises for ocamlfind, why isn't it in the official distribution ?
skchrko has quit [Read error: Connection timed out]
skchrko has joined #ocaml
<kerneis>
Anarchos: because INRIA has limited resources
<kerneis>
and they need less stuff to maintain, not more
<kerneis>
to focus on the core and important parts
Drup has joined #ocaml
<Anarchos>
kerneis seems a good explanation
<Anarchos>
i would prefer them to make an offical manual in french
<kerneis>
which would double the amount of effort, and would always get out of sync
<Anarchos>
as they are paid by french tax payers :)
q66 has joined #ocaml
<Anarchos>
kerneis official researchers can't publish all their work in english, and at the same time complain about the lack of interest for science by the population ! It is unfair
<kerneis>
I don't hear a lot of them complain
<kerneis>
I am certainly not complaining
<kerneis>
the target of scientific papers is *not* the general public
<kerneis>
and that's fine
<kerneis>
it doesn't mean researchers need not explain what they are doing too
<kerneis>
but this is a separate branch of their work
<Anarchos>
well for the manual of ocaml, it would have been kind :)
<kerneis>
I have been taking part in "Fête de la science" for several years for instance
<fds>
Even in native English-speaking countries, the general population can't read articles published by scientific researchers!
<kerneis>
no, the ocaml manual is a technical reference manual
<kerneis>
you have a number of good french books about ocaml
<kerneis>
some of them written (arguably a few years ago) by the very people who developped the compiler
<Anarchos>
ok i stop whining :)
<Anarchos>
i was a great fan of 'Le langage caml' it taught me a lot !
<kerneis>
yeah, good book
<kerneis>
Développement d’applications was nice too
<kerneis>
even if some parts are deprecated
<Anarchos>
kerneis LLC goes from a lambda calculus language to its exectution on a picoprocessor :)
<Anarchos>
"LE Langage Caml" is available on internet, being out of print :)
<adrien_oww>
wrote on the wrong channel
<adrien_oww>
12:02 < adrien_oww> the official distribution is a compiler
<adrien_oww>
12:03 < adrien_oww> not a library-locating tool
<adrien_oww>
12:03 < adrien_oww> camlp4, ocamlbuild, ocamldoc, ... are slowly being separated from the com
<Anarchos>
and how to make ocamlfind locate my dypgen installation ?
<kerneis>
Anarchos: write a META file
<adrien_oww>
find /lib/ocaml -name META | grep dyp
<adrien_oww>
returns anything?
<adrien_oww>
f00d-time
<kerneis>
then: ocamlfind install META *.{cma,cmxa,cmxs,cmo,cmi,mli}
<kerneis>
or something like that
<Anarchos>
kerneis just let me test
<Anarchos>
"/boot/common/lib/ocaml/dyp/META"
Drup has quit [Ping timeout: 245 seconds]
<Anarchos>
boot/common/lib being the standard /lib dir on my platform
mcclurmc has joined #ocaml
<Anarchos>
kerneis is it a problem ?
<def-lkb>
Anarchos: nop
<Anarchos>
kerneis "the META file is missing" but i ran the command in the dyp directory where there is the META file...
mcclurmc has quit [Read error: Connection reset by peer]
<kerneis>
Anarchos: did you actually read the manual?
<kerneis>
section 9.7
<kerneis>
(page 59)
<kerneis>
(so, yes, it's a pity it doesn't use findlib by itself)
<kerneis>
(but at least it's documented, isn't it?)
<Anarchos>
yes i read, but my parser is a very simple one
<Anarchos>
so i thought i didn't have to add includes directive
chambart has joined #ocaml
tane has joined #ocaml
yacks has quit [Quit: Leaving]
chambart has quit [Ping timeout: 264 seconds]
walter|r has joined #ocaml
yacks has joined #ocaml
ben_zen has joined #ocaml
<ggole>
module C = Camlp4;; => 19,887 lines of interface O_O
<ggole>
What the hell
<Anarchos>
one line : reference to undefined global 'camlp4' :)
<ggole>
Well, load it first
<ggole>
And dynlink (which camlp4 needs)
<ggole>
#load "dynlink.cma";; #load "camlp4o.cma"
strmpnk_ has joined #ocaml
<Anarchos>
ggole impressive indeed :)
<Anarchos>
hmmm dypgen deal with ambigous grammar and menhir allows to parametrize a parser by a module... Why can't i have both ??
<pippijn>
I need to give dypgen another go
<pippijn>
because I forgot why I chose to write my own GLR parser
<Anarchos>
pippijn there is also elkhound/elsa
<pippijn>
what about elsa?
<pippijn>
and I rewrote elkhound in ocaml :)
<Anarchos>
pippijn i had to modifiy its source code because it relies on the internal C structure of an Hashtbl, which changed when using the new MURMUR-3 algorithm
<Anarchos>
pippijn is there a c++ parser in dypgen ??
<pippijn>
Anarchos: I just made one
<Anarchos>
oh ok
<pippijn>
but it doesn't use any of dypgen's ambiguity resolution mechanisms
<pippijn>
no precedence parsing
<pippijn>
it would probably be much faster with that
<pippijn>
much much faster
<pippijn>
but I'm too lazy to do it
<Anarchos>
:)
<Anarchos>
why parsing c++ code with an ocaml parser, it means you can program in ocaml ;)
ollehar has quit [Ping timeout: 240 seconds]
yacks has quit [Ping timeout: 264 seconds]
ccasin has joined #ocaml
levi has quit [Remote host closed the connection]
shinnya has joined #ocaml
<Drup>
for anyone wondering, I asked if "%apply" was working with multiple arguments : it doesn't and it makes the compiler issue a Fatalerror ... \o/
ollehar has joined #ocaml
justdit has joined #ocaml
Neros has joined #ocaml
n06rin has quit [Ping timeout: 276 seconds]
talzeus has joined #ocaml
talzeus has quit [Remote host closed the connection]
talzeus has joined #ocaml
n06rin has joined #ocaml
<adrien>
Drup: what is the number of the bug report you've created?
<adrien>
:D
turnersr has joined #ocaml
<Drup>
one for now :3
<Drup>
I will do this one as soon as I have what I'm doing right now working
<adrien>
I meant: what is its ID so I can check it :)
<adrien>
friday evening, 5:06pm
<adrien>
I'm trusting you on that one :D
cago has left #ocaml []
derek_c has quit [Quit: leaving]
cago has joined #ocaml
cago has quit [Client Quit]
mika1 has quit [Quit: Leaving.]
Neros has quit [Ping timeout: 264 seconds]
yacks has joined #ocaml
<pippijn>
Anarchos: the bison glr parser takes 0.6 seconds for that C++ file
<Anarchos>
pippijn not too disappointing for you ?
<pippijn>
and if I invert the recursion, it can parse the ambiguous thing, as well
<pippijn>
bison handles right recursion very poorly
justdit has quit [Ping timeout: 256 seconds]
<pippijn>
ah no
<pippijn>
I'm wrong
<pippijn>
it still doesn't
<pippijn>
but now the time went down from 0.6 to 0.3 seconds
justdit has joined #ocaml
chambart has joined #ocaml
<Anarchos>
Interesting : dypgen can use external data as global_data, so it is not like parametrizing whith a module like in menhir, but not far from it !
skchrko has quit [Quit: Leaving]
jayprich has joined #ocaml
<pippijn>
I'm not sure bison can do this
<pippijn>
it *is* ambiguous
<pippijn>
and bison creates hundreds of stacks
Kakadu has quit [Quit: Konversation terminated!]
demonimin has quit [Remote host closed the connection]
<pippijn>
haha, this is insane
<pippijn>
running ocamlc -annot on a dypgen generated parser..
<pippijn>
28MB and growing annot file
<pippijn>
34M
thomasga has quit [Quit: Leaving.]
Neros has joined #ocaml
thomasga has joined #ocaml
thomasga has quit [Client Quit]
demonimin has joined #ocaml
<adrien>
code generators...
<pippijn>
the largest annot file generated from my parser generator is about 3.5MB
<pippijn>
this is stupid.. it's been at it for 15 minutes
<pippijn>
oh, I think I know
<pippijn>
dypgen produces types with 100+ type parameters
AltGr has quit [Quit: Konversation terminated!]
<Anarchos>
adrien what have you against them ?
<adrien>
nothing but they definitely produce crap code
<adrien>
well, not mine
<adrien>
but look at sexplib
<adrien>
at least one year ago, its output is much bigger that it needs to be
<pippijn>
dypgen doesn't produce crap code
<n06rin>
i write a function in file test.ml. Can i open compiler an use some operator to import this file?
<n06rin>
somethink like import in python
<Anarchos>
n06rin in the toplevel you mean ?
darkf has quit [Quit: Leaving]
<n06rin>
yes
<Anarchos>
#use "file.ml"
<n06rin>
# use "test.ml" ;;
<n06rin>
Error: Unbound value use
<n06rin>
i think i'm doing it wrong
<Anarchos>
type "#use"
<Anarchos>
the # in this case is not the propmpt !
skchrko has joined #ocaml
<n06rin>
thanks at all
chambart has quit [Ping timeout: 256 seconds]
ulfdoz has joined #ocaml
<Anarchos>
n06rin you're welcome :)
n06rin has quit [Quit: Leaving.]
n06rin1 has joined #ocaml
justdit has quit [Read error: Connection reset by peer]
justdit has joined #ocaml
<adrien>
pippijn: sorry, not crap, but redundant
<pippijn>
I'm not so sure about that
<pippijn>
I didn't look at it in detail
<pippijn>
well, I always try to minimise redundancy
<adrien>
there's little incentive to factor code in the output
<pippijn>
there is for large codes
<Anarchos>
pippijn Remark: the block %mlitop { } and the trailer make possible to encapsulate the parser defi- nition in a functor.
<Anarchos>
wow cool feature !!
<pippijn>
oh
<pippijn>
good :)
<pippijn>
so just like ocamllex
<Anarchos>
yes :)
watermind has joined #ocaml
<watermind>
So I accidentally uninstalled Core (via Opam) and now cannot install it again :(
<watermind>
I get this: ocamlfind: Conflict with another package: Library dllcore_stubs.so is already installed
<Anarchos>
watermind remove it :)
<watermind>
more precisely this
<watermind>
Anarchos: it's not there!
<watermind>
I looked at ~/.opam/4.00.1/lib/stublibs/ and it is not there
<watermind>
Anarchos: is there a way to have Opam realise it is not there anymore (even though Opam should know since it was Opam that removed it)
<pippijn>
oh, all the fun you get when reinventing the wheel
<watermind>
?
stevej has joined #ocaml
<pippijn>
yet another package manager
<watermind>
yay done!
<watermind>
pippijn: right, true
<watermind>
I'm not exactly sure how I fixed this but I did
justdit has quit [Read error: Connection reset by peer]
ontologiae has quit [Ping timeout: 245 seconds]
justdit has joined #ocaml
chambart has joined #ocaml
Neros has quit [Ping timeout: 264 seconds]
ben_zen has joined #ocaml
thomasga has joined #ocaml
jayprich has quit [Ping timeout: 245 seconds]
justdit has quit [Ping timeout: 245 seconds]
thomasga has quit [Quit: Leaving.]
justdit has joined #ocaml
mcclurmc has quit [Quit: Leaving.]
<Anarchos>
watermind no idea, i don't use opam
Anarchos has quit [Quit: Vision[0.9.7-H-130604]: i've been blurred!]
justdit has quit [Ping timeout: 260 seconds]
Drup has quit [Ping timeout: 246 seconds]
justdit has joined #ocaml
jayprich has joined #ocaml
ddonna has joined #ocaml
<ddonna>
So, I've spent the past couple hours trying to get OcaIDE to compile a project that uses lablGL. The help documentation is telling me that an External files directory will be created (it hasn't been), and I can't find an alternate route. Is OcaIDE to be considered harmful?
<ddonna>
Or am I missing something charming and hilarious? (I can compile the project with no trouble using ocamlbuild.)
<companion_cube>
does anyone know how to specify options for menhir with ocamlbuild?
justdit has quit [Read error: Connection reset by peer]
gnuvince has quit [Ping timeout: 268 seconds]
<pippijn>
saml: not bad
<pippijn>
and I didn't know about emily, yet
gnuvince has joined #ocaml
gnuvince has quit [Changing host]
gnuvince has joined #ocaml
<watermind>
saml: that complain about the generic error msg seems silly, as it is generated by his generic "assert false"... or am I missing something?
<watermind>
although I do agree it would be nice if the assert generated more info than that
Drup has joined #ocaml
es917 has joined #ocaml
es917_ has joined #ocaml
es917_ has quit [Client Quit]
flx has joined #ocaml
flx is now known as Guest71787
flux has quit [Read error: Connection reset by peer]
Guest71787 is now known as flux
paddymahoney has joined #ocaml
ggole has quit []
Snark has quit [Quit: leaving]
void64 has quit [Ping timeout: 245 seconds]
void64 has joined #ocaml
tani has joined #ocaml
tane has quit [Ping timeout: 245 seconds]
levi has joined #ocaml
_andre has quit [Quit: leaving]
weie_ has joined #ocaml
weie has quit [Ping timeout: 264 seconds]
<jpdeplaix>
gasche: is it normal that ocamlbuild builds bytecode files when executing « ocamlbuild test.native » ?
<jpdeplaix>
(I'm filling a bugreport right now)
dnm has quit []
<adrien>
jpdeplaix: missing .mli?
<jpdeplaix>
adrien: yes, why ?
<adrien>
:)
<adrien>
ocamlbuild will build .cmi files
<adrien>
from .mli files it generate
<adrien>
s
<adrien>
for that it needs to build cmo files
<jpdeplaix>
for building the .cmi files or for generates the .mli ?
<jpdeplaix>
s/generates/generating/
<jpdeplaix>
because for building the .cmi, there is no need for the .cmo
<adrien>
don't remember well; it's a kwown issue with at least one PR on mantis
<adrien>
ocamlbuild's rule does it that way
<adrien>
iirc it doesn't handle having two rules to build the same file