<Carillon>
hi, has anyone here tried building a set of Qt bindings for Ocaml?
hikozaemon has joined #ocaml
ita|zzz is now known as ita
<ita>
Carillon: no
<ita>
dylan: ping
<Carillon>
ita: ;) no?
<ita>
Carillon: but if you have any news on this i am interested
<ita>
by such bindings
<ita>
later
ita has quit ["Konversation terminated!"]
fab__ has joined #ocaml
fab_ has quit [Read error: 110 (Connection timed out)]
mefalcon has joined #ocaml
<mefalcon>
hi, I'm trying to define a type, in haskell it is given as "type Behavior a = [Time] -> [a]" in Ocaml I define it as "type 'a Behavior = [DateTime] -> ['a]" ... i get a syntax error at the first '[', did i miss something?
<mefalcon>
instead of "[sometype]" do i need to do "sometype list" ?
<zmdkrbou>
yep
<zmdkrbou>
" type 'a behavior = time list -> 'a list " is the right syntax, actually
<zmdkrbou>
(capitalized identifiers are for modules and type constructors)
jcreigh has quit ["Cuius rei demonstrationem mirabilem sane detexi. Hanc marginis exiguitas non caperet."]
d-bug has joined #ocaml
Smerdyakov has quit ["Leaving"]
shekmalhen has joined #ocaml
dibalout has joined #ocaml
dvekravy_ has joined #ocaml
fab__ has quit [leguin.freenode.net irc.freenode.net]
shawn has quit [leguin.freenode.net irc.freenode.net]
mefalcon has quit [leguin.freenode.net irc.freenode.net]
dvekravy has quit [leguin.freenode.net irc.freenode.net]
Shimei has quit [leguin.freenode.net irc.freenode.net]
shawn has joined #ocaml
falconair has joined #ocaml
dibalout is now known as tristram
Shimei has joined #ocaml
fab__ has joined #ocaml
smimou has joined #ocaml
Snark has joined #ocaml
shekmalhen has quit ["leaving"]
mnemonic has joined #ocaml
<mnemonic>
hi
michael-i has joined #ocaml
michael-i has left #ocaml []
mnemonic has quit ["leaving"]
dvekravy_ is now known as dvekravy
hikozaemon has quit ["Leaving..."]
slipstream has joined #ocaml
ertai has joined #ocaml
slipstre1m-- has joined #ocaml
slipstream-- has quit [Read error: 110 (Connection timed out)]
slipstream has quit [Read error: 110 (Connection timed out)]
fab__ has quit []
ramkrsna has joined #ocaml
dylan is now known as Dylan
shawn has quit ["This computer has gone to sleep"]
clconway has joined #ocaml
<clconway>
I am on the verge of homocide... Can't make an "Unbound module" error go away even though I'm giving "-I dir"...
Smerdyakov has joined #ocaml
datrus has joined #ocaml
<clconway>
I just know I've seen this before, but I've run out of ideas. Tried dozens permutations of command line args...
<datrus>
(i'm a ocaml newbie) i try to compile a file using "ocaml -c test.ml" and i then get an error "Unbound value Manager.make" (i use the cudd.cma BDD library in test.ml). As I understant it, linking doesn't take place at this stage yet.
mwc has joined #ocaml
<clconway>
datrus, you and I are in the same boat...
<clconway>
You need to give "-I dir" arguments pointing to the directory with manager.cmo in it
<datrus>
ok, trying...
<datrus>
can't i use cudd.cma instead? i believe manager.cmo is included inside cudd.cma
<clconway>
Should work...
<datrus>
if i understant correctly .cma is like gcc's .a
<clconway>
Yeah
<clconway>
Any CIL users in here?
<datrus>
thanks compiling the cmo worked now
<clconway>
No prob
<mwc>
.cma is basically the compiled bytecode and some higher level data
<mwc>
.a is just object code
<mwc>
the difference is that ocaml is a "smart linker", whereas ld is pretty stupid
<clconway>
You probably need to do "ocamlc -I CUDD_DIR +cudd test.cmo" to link
<Smerdyakov>
clconway, I've used CIL, and others in my research group developed it.
<Smerdyakov>
clconway, are you in MSR?
<clconway>
Yeah. How'd you guess?
<Smerdyakov>
Hostname plus channel name.:)
<Smerdyakov>
Intern?
<clconway>
Yeah
<Smerdyakov>
Who's your mentor?
<clconway>
Sriram Rajamani
<datrus>
clconway: i used ocamlc -custom -cclib -lcudd_caml -o $@ test.cmo
<Smerdyakov>
Cool. Are you in India?
<clconway>
Yeah. Bangalore
<clconway>
I'm having the darnedest time linking against CIL. But I think it's cygwin's fault.
<mwc>
I remember seeing some gotchyas on Cygwin
<datrus>
i use cygwin too btw
<clconway>
Works find as long as I put my code in CIL/src/ext... Compiling separately doesn't work at all.
<clconway>
And cilly doesn't work either: Cannot open output file /cygdrive/c/DOCUME~1/t-chric/LOCALS~1/Temp/cil-PGIrCEyw.cil.c.
shekmalhen has joined #ocaml
lmbdwr has joined #ocaml
<lmbdwr>
oi
<zmdkrbou>
io
<clconway>
Smerdyakov, do you know how to instantiate a user-defined type in CIL?
<clconway>
I.e., given "typedef <blah> T", create a value of type "T" without a priori knowing the structure of <blah>
mwc has quit [Remote closed the connection]
<Smerdyakov>
You can't just use TNamed?
<clconway>
TNamed takes <blah> as an argument
<Smerdyakov>
I don't know.
clconway has left #ocaml []
ookk has joined #ocaml
shawn has joined #ocaml
shekmalhen has quit ["bone journer"]
youknow365 has joined #ocaml
<youknow365>
chwa chea
Smerdyakov has quit ["Leaving"]
Demitar_ has joined #ocaml
creichen has quit [Read error: 131 (Connection reset by peer)]
Demitar has quit [Nick collision from services.]
Demitar_ is now known as Demitar
finelemon has joined #ocaml
finelemo1 has quit [Read error: 110 (Connection timed out)]
lmbdwr has quit [Read error: 54 (Connection reset by peer)]
lmbdwr has joined #ocaml
<ookk>
is the ocaml keyword "in" the same as the one in python
<ookk>
could it be used to check if an element is in a list?
<ookk>
"if a in list" for instance
<ookk>
hmm apparently not
<buggs>
no
<youknow365>
ookk: wtf you gonna say python in herre for bitch
<buggs>
_let_ this be (=) that _in_ there
<ookk>
yeah i just thought that "in" had another use in ocaml
<ookk>
:P
lmbdwr has quit [Read error: 104 (Connection reset by peer)]
lmbdwr has joined #ocaml
ita has joined #ocaml
<ita>
Dylan: ping?
<Dylan>
ita: pong
<ita>
Dylan: have you seen the memo ?
<Dylan>
Ah
<Dylan>
I always do that via making the generated output require the .ml files.
<Dylan>
[in answer to the memo]
<ita>
Dylan: the problem is that the .mly and .mll generate .ml files
<ita>
Dylan: the order of targets in the link matters, right ?
<Dylan>
Not a problem, if deps.bk requires foo.ml, and bar.ml, and bar.ml is generated from bar.mly, it'll all work out
<Dylan>
my current idea is to express dependency and build information in some non-language format, and then write tools to generate that. :)
<Dylan>
sort of like ant, I suppose... except without XML and cruft.
<ita>
Dylan: do you run ocamldep to find the right target order ?
shawn has quit [Connection reset by peer]
<Dylan>
ita: no, the target order I specified statically
<ita>
Dylan: waf is able to infer the right target order
<Dylan>
ocamldsort can too
<Dylan>
I just was lazy. :)
<ita>
Dylan: ocamldsort ?
<Dylan>
dependency sorter for OCaml source files
<ita>
it is not distributed with ocaml, is it ?
<Dylan>
no, it's a debian package though
<ita>
Dylan: makefile-based projects use "make depend"
<Dylan>
Not ones using gnu make
<ita>
Dylan: btw, waf also has a daemon mode now (background recompilation upon file change using fam/gamin)
<Dylan>
(they don't have to, any way)
<ita>
Dylan: not the ones using gnu make, hmm ? :-)
<Dylan>
gnu make will rebuild automatically generated included files.
<Dylan>
though, that doesn't absolve it of its sins
<ita>
gnu make does not infer the right order for the targets or the link command
<Dylan>
it doesn't infer anything
<Dylan>
but if make.deps depends on foo.c and bar.c, and those files change, it'll rebuild it
<ita>
except with the depend target which is usually "depend: \n\t $(OCAMLDEP) $(INCLUDES) *.mli *.ml > $@ "
<ita>
this is the behaviour i tried to replicate
<ita>
oh well, if nobody needs it i can turn it off as well
<Dylan>
if you write:
<Dylan>
include depend
<Dylan>
then it will rebuild depend if required
<ita>
yes
lmbdwr has quit [Read error: 104 (Connection reset by peer)]