emmanuelux has quit [Remote host closed the connection]
sepp2k has quit [Remote host closed the connection]
madroach has quit [Ping timeout: 265 seconds]
madroach has joined #ocaml
hongboz has joined #ocaml
braibant has quit [Quit: Leaving.]
Xizor has quit []
braibant has joined #ocaml
<dsheets>
ocamllex syntax highlighting lost in typerex 1.0.1… but "rule" appears to be a keyword in tools/ocp-typerex-ide/emacs/typerex-tuareg.el emacs wizards?
jamii has joined #ocaml
sivoais has quit [Remote host closed the connection]
sivoais has joined #ocaml
braibant has quit [Quit: Leaving.]
jamii has quit [Ping timeout: 276 seconds]
<cacho>
anyone knows about that ocamlbuild thing?, already separated dirs to continue development but would be pretty good to fix it
hongboz has quit [Ping timeout: 248 seconds]
BiDOrD_ has joined #ocaml
BiDOrD has quit [Ping timeout: 252 seconds]
ankit9 has joined #ocaml
scri has joined #ocaml
<scri>
i just got 4.00 and i'm trying to build batteries-included
<scri>
tells me: File "_none_", line 1:
<scri>
Error: Files /usr/lib/ocaml/unix.cmxa and /usr/local/lib/ocaml/stdlib.cmxa
<scri>
make inconsistent assumptions over interface Array
<scri>
any suggestions?
<dsheets>
scri: are unix.cmxa and stdlib.cmxa both fresh after your 4.00 install?
<scri>
why wouldn't they be?
<scri>
do i need to do a complete removal of 3.12?
<dsheets>
scri: it would appear that one was compiled against a new version of Array and one against an old version
<dsheets>
you don't necessarily need to remove 3.12 but you do need to make sure that your linking uses a consistent set of compiled modules
<dsheets>
are you using ocamlfind? examine ocamlfind printconf
<dsheets>
then, examine your ld.conf file
<scri>
The standard library is assumed to reside in:
<scri>
/usr/lib/ocaml
<scri>
<scri>
includes lib and local/lib in the search path
<scri>
ok, what are the stublibs? there's two here
<scri>
but nothing in local/lib/ocaml/3.12.0/stublibs/
<dsheets>
stublibs are for shared libraries
<dsheets>
what process did you use to install 4.00 on your 3.12 system?
<scri>
built from source
<scri>
i see the libraries in /lib, but /local/lib is empty - would it be simpler to just remove 3.12? i don't need it
<dsheets>
if /local/lib is empty, how does '/usr/local/lib/ocaml/stdlib.cmxa' exist?
<scri>
ah, i mean, the stublibs directory is empty
<scri>
sorry
<dsheets>
stublibs is a red herring
<dsheets>
your link path includes compiled modules built against different Array modules… perhaps the 4.00 build installer modified your existing findlib config?
* dsheets
does not have 4.00 yet
<scri>
hm, if i just want to remove the old version, would it work to remove the /local/lib directory?
<dsheets>
Probably but then you will have references to the local/lib directory in your configs that will always fail. I believe you could simply modify your link settings (in ocamlfind or your custom build system) to not search the 3.12 install
Derander has quit [Ping timeout: 244 seconds]
Derander has joined #ocaml
K_F has quit [Ping timeout: 240 seconds]
K_F has joined #ocaml
<rgrinberg>
how to change oasis to use ocamlopt instead of ocamlc?
ankit9 has quit [Quit: Leaving]
<rgrinberg>
nvm. CompiledObject: native does it.
rgrinberg has quit [Ping timeout: 246 seconds]
<scri>
ok that issue is resolved, ty
<scri>
but trying to build batteries again, and getting "unbound module Toploop"
sepp2k has joined #ocaml
<adrien>
ok, the forge is nice, I love it
<adrien>
but
<adrien>
it's not visible in search engines results
<adrien>
or not always as high as it should
<adrien>
and there are quite often the old pages for the projects that have moved
sgnb has quit [Read error: Connection reset by peer]
sgnb has joined #ocaml
Anarchos has joined #ocaml
Snark has joined #ocaml
<madroach>
Hi, is there somewhere documented which +compiler-libs should be available on a bytecode-only architecture?
silver has joined #ocaml
companion_cube has joined #ocaml
eikke has joined #ocaml
<adrien>
We are currently looking for two one-year postdocs at INRIA Saclay[0],
<adrien>
to work on improving OCaml memory management, particularly in the
<adrien>
context of multicore systems.
<adrien>
:P
<companion_cube>
oh dear
<companion_cube>
in short, they want to give a try to concurrent GC
<adrien>
seems there's a strong corporate demand for it so we'll see
<adrien>
but oc4mc already has working stuff
<companion_cube>
then, merge JoCaml in mainstream... ;)
<adrien>
heheh
<rixed>
adrien: wasn't already a concurrent GC in an old version of ocaml? which was abandoned because bugged?
<adrien>
not of ocaml: that was in caml super light
<rixed>
IIRC this was proved correct, but was bugged nonetheless :)
<rixed>
which shows how hard this kind of thing is
hongboz has quit [Remote host closed the connection]
hongboz has joined #ocaml
hongboz has quit [Ping timeout: 272 seconds]
<thomasga>
adrien: oc4mc was a nice experiment but that's it
<thomasga>
it's not maintain (and it is hard to maintain)
<thomasga>
maintained
<thomasga>
and the goal was not to have a usable compiler, just to understand the problems
<thomasga>
(usable runtime)
djcoin has joined #ocaml
<Anarchos>
adrien maybe the doligez concurrent GC could be proven in coq instead of TLA, and then the code could be extracted ?
<companion_cube>
coq-generated code may be too slow for something as low level as a GC, I fear
sepp2k has quit [Ping timeout: 246 seconds]
<adrien>
I can't remember who mentionned that coq-extracted code was really too slow usually, not long ago, on the caml-list
<adrien>
thomasga: ah, ok; I had thought it was at least not too hard to maintain
<Ptival>
companion_cube: the author of JOCaml will be integrated in the OCaml team soon actually :D
<Ptival>
not that it implies anything on the integration you mentioned!
cdidd has joined #ocaml
<companion_cube>
adrien: but the GC is really performance critical, isn't it? I doubt it is written in a functional style, more is
<thomasga>
(the GC is written in C usually)
<thomasga>
(as you don't want to garbage collect the GC code as well)
<adrien>
companion_cube: oh, yeah, agreed; plus a GC in functional style which would use a GC for its allocations, would be funny
<adrien>
ATS? :P
<companion_cube>
:>
<Ptival>
s/adrien/Anarchos/ in companion_cube's message (meh, triple HL)
<companion_cube>
adrien: come on, that's where the "concurrent" is important ! you write a GC to GC the proven GC's garbage ;)
<adrien>
hahaha :P
* orbitz
wants a GC that is just a Coq code extraction :)
<companion_cube>
let's have coq generate low level C
<companion_cube>
(maybe from compcert? :))
sepp2k has joined #ocaml
<f[x]>
thomasga, so the current settled approach is one runtime per thread instead?
<Anarchos>
companion_cube i mean to say : extract code from coq, and then translate it in C.
<Anarchos>
adrien could you get a look on anubis ?
<adrien>
Anarchos: whihc is?
<Anarchos>
adrien a language made by Alain Prouté, based on category theory.
<Anarchos>
adrien and it was caml special light, not super light ;)
Yoric has joined #ocaml
scri has quit [Ping timeout: 246 seconds]
Yoric has quit [Remote host closed the connection]
Yoric has joined #ocaml
ankit9 has quit [Quit: Leaving]
<madroach>
why is ocamlbuild installed with .native suffix instead of .opt?
<adrien>
because ".opt" is misnamed
Anarchos has quit [Quit: Vision[0.9.7-H-090423]: i've been blurred!]
emmanuelux has joined #ocaml
<madroach>
ok. My problem is all executables installed by ocaml are named according to what they are. Except ocamlbuild. ocamlbuild without any suffix is bytecode on sparc64, but nativecode on amd64.
<madroach>
this is painfull when writting portable package build scripts.
thomasga has quit [Quit: Leaving.]
_andre has joined #ocaml
jonafan has quit [Ping timeout: 245 seconds]
jonafan has joined #ocaml
Yoric has quit [Ping timeout: 265 seconds]
<adrien>
madroach: ocamlbuild has plugins; I guess native plugins aren't available on sparc64
<adrien>
madroach: and check if ocamlbuild.byte exists; also, by default, consider that you only have bytecode
emmanuelux has quit [Ping timeout: 244 seconds]
Yoric has joined #ocaml
eikke has quit [Ping timeout: 272 seconds]
Cyanure has quit [Remote host closed the connection]
Cyanure has joined #ocaml
gnuvince has quit [Ping timeout: 248 seconds]
Anarchos has joined #ocaml
eni has quit [Ping timeout: 268 seconds]
eni has joined #ocaml
eikke has joined #ocaml
Progster has joined #ocaml
<_andre>
is there a reason i'm getting "‘caml__temp_val’ undeclared" errors in a C stub even though i have included caml/mlvalues.h?
<_andre>
i've only seen this error when i forgot the header
thomasga has joined #ocaml
gnuvince has joined #ocaml
cacho has quit [Ping timeout: 246 seconds]
<adrien>
_andre: show your code
<madroach>
adrien: thanks for the info. I solved it by simply symlinking ocamlbuild to either ocamlbuild.native or ocamlbuild.byte.
<_andre>
adrien: never mind... i had a variable called "value" :|
<adrien>
:D
ankit9 has joined #ocaml
cacho has joined #ocaml
emmanuelux has joined #ocaml
emmanuelux has quit [Read error: Operation timed out]
eikke has quit [Ping timeout: 246 seconds]
Progster has quit [Ping timeout: 246 seconds]
Anarchos has quit [Quit: Vision[0.9.7-H-090423]: i've been blurred!]
ftrvxmtrx has quit [Quit: Leaving]
ftrvxmtrx has joined #ocaml
mika1 has quit [Quit: Leaving.]
hcarty has quit [Remote host closed the connection]
sabayonuser2 has joined #ocaml
sabayonuser2 is now known as Mnabil
hcarty has joined #ocaml
Mnabil has quit [Client Quit]
Mnabil has joined #ocaml
<hcarty>
gildor: When you're around the #ocaml topic could use a refresh for the 4.00.0 release.
eikke has joined #ocaml
<_andre>
is it possible to use C sources in an Executable in oasis, without having a separate Library for the stub?
<f[x]>
_andre, yes
<_andre>
f[x]: i'm trying this but compilation fails saying it can't find dlltest_stubs.so (my program is test.ml)
<_andre>
the C code must be named test_stubs.c in that case?
<semperos>
that's actually the same definition I found when perusing OCaml's source distro, in the ocamlbuild folder
<semperos>
learning OCaml; per usual, the ecosystem (build tools, stdlib(s)) has been harder to navigate than the language itself :)
<thelema>
semperos: ask lots of questions
<semperos>
will do
<semperos>
I'm trying out building my first project using ocamlbuild via OASIS, so far so good
avsm has quit [Quit: Leaving.]
ankit9 has quit [Ping timeout: 246 seconds]
<thelema>
great. Oasis has a lot of potential for improving the ocaml ecosystem
K_F has joined #ocaml
avsm has joined #ocaml
ankit9 has joined #ocaml
Mnabil has quit [Ping timeout: 272 seconds]
ftrvxmtrx has joined #ocaml
ankit9 has quit [Quit: Leaving]
djcoin has quit [Quit: WeeChat 0.3.2]
Obfuscate has quit [Changing host]
Obfuscate has joined #ocaml
Guest34974 has quit [Changing host]
Guest34974 has joined #ocaml
Guest34974 is now known as henux
<wmeyer`>
hcarty: So I managed to install OCaml on AC100. However, there were some problems, and I could not fully utilize ocamlbrew. First thing, lack of memory made it swap heavily - until actually I ran in pure text console. Second, gcc had a bug that caused hang during compilation of one file. So for this issues, I think ocamlbrew could have a --continue option, or option of installing everything apart from OCaml. The installation
<adrien>
they look a bit like "oh, it's awesome, it's perfect, it's gorgeous", btu?
<adrien>
s/btu/but/
<thelema>
but using sqrt and modulus in something that's supposed to be fast...
<thelema>
I guess if m is a even power of 2
<adrien>
well, they're supposed to have awesome algorithmic complexity
<thelema>
I've never used them
<adrien>
but I've first heard about them recently and I'm wondering if they're not quite more complex to implement and thus not better
<wmeyer`>
"However, for small trees the overhead associated with vEB trees is enormous: on the order of 2m/2. This is one reason why they are not popular in practice. One way of addressing this limitation is to use only a fixed number of bits per level, which results in a trie. Other structures, including y-fast tries and x-fast tries have been proposed that have comparable update and query times but use only O(n) or O(n log M) space w
<wmeyer`>
the number of elements stored in the data structure."
Cyanure has quit [Read error: Connection reset by peer]
Arsenik has joined #ocaml
<adrien>
I had read that but back then I had skipped over something else
<adrien>
log m = k, and k is an integer
<adrien>
but maybe it's fine in practice
eikke has joined #ocaml
<wmeyer`>
you can't lower the bounds of the complexity of data structure - therefore you will pay in one or other way. Usualy, it means that data structure will be unstable for some sets of data
<wmeyer`>
it affects also how complex the data structure is to implement. Hash tables are best example, they are easy to implement, but really they are? Right and efficient hash function is so trivial
<wmeyer`>
not so trivial*
<adrien>
yeah, I was thinking it couldn't be better all-round; but if noone here has ever come over any implementation close to 40 years after the publication, there might be a reason
gnuvince has quit [Quit: Remember when men were men and regular expressions recognized regular languages?]
gnuvince has joined #ocaml
Yoric has joined #ocaml
Yoric has quit [Remote host closed the connection]
Yoric has joined #ocaml
cacho has quit [Ping timeout: 240 seconds]
eni has quit [Ping timeout: 240 seconds]
Mnabil has joined #ocaml
ftrvxmtrx has quit [Ping timeout: 260 seconds]
Arsenik has quit [Remote host closed the connection]
avsm has quit [Quit: Leaving.]
eikke has quit [Ping timeout: 252 seconds]
<testcocoon>
does anybody have some experience with ocamldebug on Windows? it seams that Ctrl-C does not not the application but quits ocamldebug and with the version 4.0 it try to connect to a local server
_andre has quit [Quit: leaving]
sepp2k has quit [Remote host closed the connection]