<wysek>
how to transform this command into ocamlbuild?
<wysek>
anyone here?:)
Spiwack has joined #ocaml
love-pingoo has joined #ocaml
komar_ has joined #ocaml
komar__ has joined #ocaml
<kaustuv>
wysek: add use_unix and use_sem tags to semtest.ml using the _tags file, and then write a myocamlbuild.ml that adds the -L .. and -lpthread options for tag use_sem
<wysek>
I've found out a better way than ocamlopt:
<wysek>
ocamlfind opt -package sem -cclib -lpthread sem.cmxa unix.cmxa semtest.ml
<Alpounet>
It's at least 13pm @ Manhattan, right ?
<Alpounet>
Pfft, they're probably eating and driking some wine, laughing about how they'll announce JSSP projects.
j_lenorm has joined #ocaml
<olegfink>
Alpounet: you know, there's rss...
kaustuv has joined #ocaml
<Alpounet>
olegfink, heh, F5-ing is fine in my case. I'm not concentrated on anything else for the moment, so ...
* j_lenorm
is a happier person since he stopped using gtk
<kaustuv>
Alpounet: were you selected? Do you know yet?
<Alpounet>
kaustuv, I don't know yet.
<Alpounet>
JSSP FAQ says awards will be announced today.
<Alpounet>
But still nothing on their page... But it's only 13pm there I think.
<kaustuv>
Wow, SML and Haskell proposals also this year?
<j_lenorm>
yeah
<j_lenorm>
I say we get a contigency in google's summer of code for next year
<Alpounet>
kaustuv, yep, but OCaml has priority.
<Alpounet>
+"a higher"
<j_lenorm>
err
<j_lenorm>
what does rss do?
<Alpounet>
j_lenorm, it is integrable in your bookmarks or so (Firefox, ...), so that you don't have to load the whole page to know if there's a new post on a blog, etc.
<j_lenorm>
cool
<j_lenorm>
any comments? I don't have an account so I can't check
deech has quit [Read error: 104 (Connection reset by peer)]
<flux>
alpounet, iirc last time it took ages to publish the winning projects. never mind how long it took to publish some real information on them as well..
<Alpounet>
if I'm selected, for informations about mine, you can ask here
<Alpounet>
I plan to publish many stuffs about it on planet ocaml, about what has been done, what is being done, etc (still if I'm selected)
<flux>
alpounet, what was the project you proposed?
prime2 has joined #ocaml
<flux>
(an executive summary ;))
<Alpounet>
An alternative code generation for OCaml's compiler, based on HLVM/LLVM, handling well parallelization and numeric computations (as opposed to the current implementation which is much better for symbolic computations
Associat0r has joined #ocaml
<Spiwack>
Which would create life-long nightmares for people doing projects including a half of symbolic computation and a half of numeric computation
<Spiwack>
Ça a pas l'air d'être ça
<Spiwack>
.(
<Spiwack>
Je vais quand même pas devoir partir en weekend sans avoir résolu ça :(
<Spiwack>
ops
<Spiwack>
wrong window every one :p
<Spiwack>
Sorry for the interesting spam :p
<Alpounet>
Spiwack, interesting for people like me who also speak French :-)
<Alpounet>
Spiwack, then it would be interesting to introduce optimizations for getting a better compromise
<Alpounet>
but it's not the aim of my project... And moreover, that would be interesting for scientific computing projects, with many numerical computations.
_andre has quit ["leaving"]
<Spiwack>
I'm not sure it was to be taken seriously :p
<Alpounet>
if it was not, I had nothing special to answer. If it was, I answered this way :-p
jzmer has joined #ocaml
<jzmer>
is there any good linear algebra, especially symbolic linear algebra libraries or modules around for ocaml?
<Alpounet>
have you searched in the (O)Caml hump ?
<jzmer>
well i did find lacaml
<jzmer>
but the problem is that lacaml does not support symbolic
<jzmer>
and lacaml does not have any docs i could locate through googling
<jzmer>
and that is the only item u could find in caml hump when u typed in "linear algebra"
Spiwack has quit ["Leaving"]
<Alpounet>
ok jzmer
<Alpounet>
Any "major" project in OCaml is added to the Caml hump
<jzmer>
Alpounet: u mean google "major" in caml hump?
<Alpounet>
but you should find some doc in the lacaml release
<jzmer>
Alpounet: currently im using my distro's ocaml-lacaml package, which does not even have any example
<Alpounet>
jzmer, are you sure ?
<jzmer>
Alpounet: but i could also try the package from caml hump, which might have something more than those provided in the distro package
<Alpounet>
jzmer, indeed, there isn't any doc in the release archive... But there are examples in the "examples" dir, and the code is commented, it can be of help ;-)
<Alpounet>
Moreover, you can contact lacaml's author IMO.
<jzmer>
Alpounet: thanks, checked the package at caml hump, does have some examples, just wonder my distro didnt include those . . .
<Alpounet>
jzmer, it should... as it may just "transform" the archive to fit your system's hierarchy?.
<Alpounet>
s/?//
<jzmer>
well on my distro th e package is split into ocaml-lacaml and ocaml-lacaml-devel . and i did a rpm -ql with grep -i example , and it turns out nothing
<jzmer>
and it also seems to be somewhat oudated . . .
<Alpounet>
ok
<Alpounet>
jzmer, you should rather use the release archive published on ocaml.info
<jzmer>
anyway, u could rarely see outdated packages in fedora, i would probably contact the pkg maintainer
<Alpounet>
Debian is better for OCaml IMO :-)
<jzmer>
Alpounet: thanks, i just know hump before
<jzmer>
dunno whether debian apply optimization flags and safety stacks and other feature flags to compiler for ocaml though
<jzmer>
in fedora they have all those flags enable and the build log is visible
<mrvn>
check on buildd.debian.org
prime2 has quit ["leaving"]
<jzmer>
mrvn: thanks, checked the debian ocaml packages, but they seem to use less flags though
<bluestorm>
hlab: probably means you've been messing with the toplevel
<bluestorm>
two different types with the same name defined at different times
<hlab>
what does that mean? :P
<hlab>
ahh!
<hlab>
gotcha
<palomer>
hrmph
<palomer>
I'm writing a toolkit
<palomer>
that depends on an external library called foo.cmxa
<palomer>
erm...
<palomer>
I better include foo.cmxa into my library I guess
kaustuv has quit [Remote closed the connection]
kaustuv has joined #ocaml
deech has quit [Read error: 104 (Connection reset by peer)]
deech has joined #ocaml
Yoric[DT] has joined #ocaml
ttamttam has joined #ocaml
<mfp>
bluestorm, Alpounet: OK, pushing it in a while then
<mfp>
(I've already rebased it against latest master)
<Alpounet>
mfp, I'll just wait it's in master to create tests on monad stuffs
<mfp>
actually, s/in a while/right now/
<mfp>
so you have no excuse }:-)
<Alpounet>
Haha.
<Alpounet>
I'll write them tonight, still waiting for JSSP anouncement :-p
<mfp>
pushed
monadic_kid has joined #ocaml
psnively has quit []
<Alpounet>
mfp, cheers ;-)
<bluestorm>
i've tried to adapt my camlp4 testing to the OUnit framework but's it's really quite difficult
<bluestorm>
as it is very sensitive to the compilation options, and I haven't found the way to build then painlessly with ocamlbuild yet
<bluestorm>
(it will probably need some ocamlbuild plugin)
ulfdoz_ has quit [SendQ exceeded]
ttamttam has left #ocaml []
_andre has quit ["Lost terminal"]
_andre has joined #ocaml
_andre has quit [Client Quit]
Ariens_Hyperion has joined #ocaml
komar_ has quit [Read error: 104 (Connection reset by peer)]
<hcarty>
What is the proper way to use a polymorphic variant in the OCaml C ffi? Is Int_val(hash_variant(foo)) appropriate?
thelema has joined #ocaml
<mrvn>
doesn't hash_variant() return a C int?
monadic_kid has quit ["Leaving"]
jamii has quit [Read error: 60 (Operation timed out)]
<Alpounet>
good night all... let's see tomorrow for JSSPs
Alpounet has left #ocaml []
mlbot has quit [Remote closed the connection]
hkBst has quit [Read error: 104 (Connection reset by peer)]
Ariens_Hyperion has left #ocaml []
bluestorm has quit [Remote closed the connection]
jamii has joined #ocaml
jamii has quit [Remote closed the connection]
willb has quit ["Leaving"]
<hcarty>
mrvn: It returns value
<hcarty>
But it looks like it's a C int returned as a value
<hcarty>
So I'm not quite certain what it is supposed to do
<hcarty>
I want to compare hash_variant("foo") with (value variant_foo) and I'm not sure if (hash_variant("foo") == variant_foo) is the right thing to do
<hcarty>
It seems to work
<hcarty>
I don't want it to introduce other bugs later on though
Ariens_Hyperion has joined #ocaml
singintime has joined #ocaml
<singintime>
hello world
singintime has left #ocaml []
deech has quit [Read error: 104 (Connection reset by peer)]
deech has joined #ocaml
<palomer>
hrmph
CoryDambach has joined #ocaml
Yoric[DT] has quit ["Ex-Chat"]
jeanbon has quit [Read error: 104 (Connection reset by peer)]
<palomer>
anyone know of any good auto completion algorithms?