chickenzilla has quit ["J'y trouve un goût d'pomme."]
|jedai| has joined #ocaml
xah_lee has joined #ocaml
<xah_lee>
does anyone use hol here?
<xah_lee>
seems to have a build problem with ocaml 3.11.x
<xah_lee>
but works with 3.08
Alpounet has quit ["Ex-Chat"]
ched_ has quit [Remote closed the connection]
Ched has joined #ocaml
_dartelin has joined #ocaml
_dartelin has quit [Client Quit]
rwmjones_lptp has quit ["This computer has gone to sleep"]
itewsh has quit ["There are only 10 kinds of people: those who understand binary and those who don't"]
dartelin has quit [Read error: 110 (Connection timed out)]
seafood has quit []
xah_lee has quit ["banned in #emacs by johnsu01 (john sullivan)"]
<Camarade_Tux>
can anyone guess why this doesn't work ? (I finally found out)
<Camarade_Tux>
let genome_pid = create_process_env "flam3-genome" [||] base_env stdin genome_file genome_err_file in
<Camarade_Tux>
that's something that always catchs me
vuln has joined #ocaml
slash_ has quit ["Lost terminal"]
Demitar has joined #ocaml
verte has joined #ocaml
Camarade_Tux has quit ["Leaving"]
angerman has joined #ocaml
kaustuv has joined #ocaml
<thelema>
drat. if camarade_tux comes back, let him know that he has to pass the name of the program again as the first entry in the argv array.
angerman has quit []
angerman has joined #ocaml
alexyk has joined #ocaml
vuln has quit ["leaving"]
wsmith84 has joined #ocaml
ched_ has joined #ocaml
Ched has quit [Read error: 110 (Connection timed out)]
wsmith84 has quit ["Leaving"]
wsmith84 has joined #ocaml
<wsmith84>
quick question
<wsmith84>
newbie question
<m3ga>
wsmith84: don't ask to ask, just ask
<wsmith84>
If you create an array of floats, will they be allocated as a single contiguous array in memory, or will you obtain an array of 3 pointers, pointing to new float objects?
<wsmith84>
(A comment about strings in the docs makes me think it might do an array of ptrs)
<m3ga>
floats contained in arrays are unboxed
<wsmith84>
Can I conclude arrays are always unboxed then?
<wsmith84>
(or more precisely, that they always contain unboxed values)
<m3ga>
actually what I said is only true of the native compiler
<m3ga>
no, if you have an array of eg records, they will be unboxed
<wsmith84>
(thx m3ga, that's the only one I'm interested in)
<wsmith84>
is there any way to verify what the compiler did, apart from generating intermediate code?
<m3ga>
sorry i meant boxed for records, unboxed fro float
<m3ga>
not that i know of
<wsmith84>
I got that.
<wsmith84>
... thought technically it wouldn't be hard to make the records unboxed (they woudl just contain ptrs though, I'm guessing)
<wsmith84>
hmmm
<wsmith84>
m3ga: How about a record? Is it really implemented like a C struct?
<wsmith84>
Meaning, contiguous in one chunk of memory, or an array of ptrs?
<m3ga>
the struct itself is, so floats and ints in the struct will be unboxed, but lists/arrays/records etc that are part of a struct will be boxed
jeddhaberstro has quit []
<wsmith84>
thx m3ga
alexyk has quit []
gaze__ has quit []
<thelema>
wsmith84: although everything not representable as an int is boxed, floats have two special exceptions: arrays of floats and records (not tuples) of only floats.
<wsmith84>
Optimizations for common cases, ad-hoc I'm guessing.
<wsmith84>
thelema: just wondering, what's your tool for looking up documentation on a specific function? do you use tags?
<wsmith84>
or just source?
<wsmith84>
For e-lisp, there is pretty nice Emacs integration, e.g. prints args def in minibuffer. Would it be doable to add somthing like that for Caml? There's no introspection, right?
<wsmith84>
(sorry, was unclear, i meant the minibuffer shows help in the minibuffer when the cursor is over a lisp function)
love-pingoo has quit ["Connection reset by pear"]
<wsmith84>
If I wanted to represent a number internally as a decimal number (with fractions), e.g. for accounting purposes, is there a type in the std lib you would recommend I use? (other than using a 10**8 multiple of int or something like that)
<wsmith84>
e.g. something like Python's "decimal" module
<Alpounet>
Is there a way to tell ocamlc (installed with godi) where he should look for .cma's e.g ?
seafood has quit []
<flux>
-I ?
<Alpounet>
Am I obliged to copy them in $GODI/lib/ocaml/ ?
<flux>
you can use any paths with -I?
<flux>
I've just installed my own libraries with ocamlfind too, and let it handle the issue
<Alpounet>
-I add also the given directories for looking for CMAs ?
<flux>
well, the manual page doesn't say that, but I'm in the belief that it does, yes
<Alpounet>
ok
<Alpounet>
it seems to be okay
<Alpounet>
another error to fix now :p
<th5>
Does anyone else use ocamlspotter?
<th5>
I started using it this week and its great (well, great as in kind of thing java guys have had in eclipse for ages). One thing I don't like is that it has to patch the compiler. Doesn't seem to be packaged for any distributions.
<flux>
I haven't used - possibly for the reason you just said..
<flux>
iirc there are plans to enhance the builtin capabilities of the ocamlc to produce information in a level required by ocamlspotter
<flux>
but I'm not sure if the plans go far enough
<th5>
hmm
<th5>
yeah - right now i have a separate "sandbox" folder which the patched compiler and a few different libraries
<th5>
but its great for jumping into the source of a big project
<th5>
i think i remember xavier leroy commenting about extending .annot files but havent heard anything since
<th5>
(on the caml list)
willb has quit [Read error: 110 (Connection timed out)]
<Alpounet>
When installing lablgtk with godi, I get : > ===> Configuring for conf-gtk1-1
<Alpounet>
> Exception: Failure "Script gtk-config not found".
<Alpounet>
<Alpounet>
any solution ?
<flux>
install gtk1 -related development packages to your operating system?
<Alpounet>
I think they are
<Alpounet>
:/
<Alpounet>
now I get (for lablgl) : > configure: error: gl enforced but no support found
<flux>
well, gtk-config should come with it.. unless they also switched to pkg-config, I don't remember. (gtk2 uses pkg-config)
<flux>
install gl and glu-related development libraries
<Alpounet>
that's weird... I should have them.
<Alpounet>
I've installed them and still get this error.
<Alpounet>
Oh, fixed now. Great !
willb has joined #ocaml
kaustuv has joined #ocaml
sporkmonger has joined #ocaml
kaustuv has quit ["ERC Version 5.3 (IRC client for Emacs)"]
love-pingoo has joined #ocaml
wsmith84 has joined #ocaml
oriba has joined #ocaml
itewsh has joined #ocaml
ttamttam has left #ocaml []
Stefan_vK has quit [Read error: 104 (Connection reset by peer)]
Stefan_vK has joined #ocaml
slash_ has joined #ocaml
kaustuv has joined #ocaml
s4tan has quit []
Yoric[DT] has joined #ocaml
gaze__ has joined #ocaml
th5 has quit []
<Yoric[DT]>
hi
Snark has joined #ocaml
<Alpounet>
hi
infoe|wk has joined #ocaml
angerman has quit []
Asmadeus_ has joined #ocaml
wsmith84 has quit [Read error: 110 (Connection timed out)]
Asmadeus has quit [Read error: 104 (Connection reset by peer)]
Asmadeus has joined #ocaml
Asmadeus_ has quit [Nick collision from services.]
<kaustuv>
In ocamlbuild, how do I get a .inferred.mli target to use the same tags as when byte compiling the equivalent .ml file?
kaustuv_ has joined #ocaml
kaustuv has left #ocaml []
kaustuv_ is now known as kaustuv
<svenl_>
R/win 6
_zack has quit ["Leaving."]
jeddhaberstro has joined #ocaml
Amorphous has quit [Read error: 110 (Connection timed out)]
Amorphous has joined #ocaml
anryx has joined #ocaml
<kaustuv>
I have some non-ocamlbuild-managed code in [../libs/stuff] and I want to add it to my search path. Apparently I can't say [ocamlbuild -I ../libs/stuff] because [-I] only takes implicit paths. How do I coax ocamlbuild to add [-I ../libs/suff] (or [-I ../../libs/stuff] as the case is) to commands?
<mrvn>
What makes a path implicit?
<kaustuv>
It must be relative and cannot begin with . or ..
<kaustuv>
./ or ../, rather
<mrvn>
that makes no sense
<mrvn>
why not -I /abs/path/libs/stuff ?
<kaustuv>
For some reason ocamlbuild does not accept absolute paths for [-I].
<kaustuv>
(I am sure there is a good reason, but it isn't documented anywhere, not even in the ocamlbuild source.)
OChameau has quit [Read error: 113 (No route to host)]
<mrvn>
can't be. How else do you include /usr/lib/ocaml/3.11?
<flux>
mrvn, what do you mean 'how else'? that's the root for -I +foo-stuff?
<kaustuv>
Yes, +whatever is considered to be an "implicit path" according the rules.
<mrvn>
ok, bad example.
<flux>
but indeed, there must be some mechanism, unless that's the part godi's patched
<flux>
because ocamlfind query str -> /opt/stow/godi/lib/ocaml/std-lib and ocamlfind query extlib -> /opt/stow/godi/lib/ocaml/pkg-lib/extlib
<kaustuv>
I think if I were to use ocamlbuild with ocamlfind I wouldn't have this problem because ocamlfind would take care of the magic. Unfortunately, I cannot depend on ocamlfind.
ttamttam has joined #ocaml
<flux>
why not?
<kaustuv>
Instructions from higher ups, out of my control.
<flux>
well, ocamlfind can tell its incantations
<flux>
with some verbosity flag
<flux>
if ocamlfind can do you, ocamlc can too :)
<flux>
ops, do IT :-)
<kaustuv>
I can run [ocamlbuild -ocamlc "ocamlc -I ../../libs/stuff"], of course, but that's so far been too ugly to seriously entertain. But if that's what's needed, that's what I'll end up doing.
<flux>
perhaps the mailing list can give you advice
<flux>
or possibly encourage you to submit a bug report :)
<Yoric[DT]>
You are in the middle of a conflicted merge.
<mrvn>
Isn't git fun? So much easier than say svn. :)
<Yoric[DT]>
Ok, resolved.
<Yoric[DT]>
:)
<gildor>
s/svn/darcs/
* mfp
much prefers git's UI to darcs' nowadays, in fact (and he used darcs for a couple years before using git)
<mfp>
Yoric[DT]: what did you have to do, just resolve manually, git add & commit, or something else?
<Yoric[DT]>
A moment, I'm not out of trouble yet.
* gildor
use git and darcs and for day-to-day dev, on a single project (no branch) prefer darcs over git
<gildor>
mfp: but this is just a matter of taste
<mrvn>
I always run into problems with git.
<Yoric[DT]>
thelema: ping
<mrvn>
I had enough cases where I had to clone a new git repository, diff the 2 working dirs and import the patch in the new one.
<mfp>
gildor: well, there's a part that belongs to taste, and then there are objective problems. I ran into the exponential merge issue even for small repositories (good thing it's mostly fixed now). I still have problems with darcs when sending patches (encoding and stuff). I recently had to get a repository a couple times and reapply patches manually.
<mfp>
regarding taste & UI, I prefer git's pager + colorization, and abhor the way to refer to patches in darcs
_zack has joined #ocaml
<gildor>
mfp: never ran into this problem for now, maybe because I never tried the case you have encountered. git is a good tool but a little bit cryptic to beginner
<gildor>
mfp: that is my principal concern. For people coming from svn background, git is hard to understand, whereas darcs is a little less cryptic
* Yoric[DT]
will take a break.
* Alpounet
is compiling LLVM and HLVM... which we can call a break
<gildor>
mfp: with time, I have no doubt that git is superior in every aspect
<mfp>
gildor: I wasn't doing anything special at all, just sending a few patches from my local ocsigen branch. I was told the patches generated by darcs were corrupted :|
<mfp>
gildor: yes, darcs is probably easier to learn, and still easier to use than git in some ways, but the UI has remained relatively stale for too long, and I feel it gets in the way sometimes
<mfp>
this is getting offtopic though
<gildor>
indeed
* gildor
go back to work on oasis.ml and ocaml-autobuild
<Yoric[DT]>
How do I switch branch?
<Yoric[DT]>
Ok, found out.
* Yoric[DT]
has no clue what's going on with the git.
<Yoric[DT]>
Ok, everything seems fixed.
m3ga has quit [Read error: 113 (No route to host)]
kaustuv has joined #ocaml
jeddhaberstro has quit []
<mrvn>
What git needs is a nice frontend. :)
sporkmonger has quit []
<kaustuv>
It has one. It's called hg convert.
sporkmonger has joined #ocaml
m3ga has joined #ocaml
ttamttam has left #ocaml []
sporkmonger has quit [Client Quit]
thermoplyae has joined #ocaml
<kmkaplan>
Hello, anybody know how to easily use caml4p?
<kmkaplan>
I would like to evaluate an expression if and only if it’s compiling on ocaml 3.11 or later.
m3ga_ has joined #ocaml
<kaustuv>
Haha! Love the new batteries banner.
<kaustuv>
Though one of them should surely point the other way?