<kaustuv>
Yeah, a self extracting archive would be great, except I need to first ask the user if they have Cygwin installed and help them install it if they don't.
<Camarade_Tux>
depending on your needs, you could also extract cygwin that way
<Camarade_Tux>
I mean, bypass the installer and just extract
<Camarade_Tux>
which is going to be broken because of a few paths actually
<kaustuv>
Right, but cygwin gets very schizophrenic if there is more than one install of cygwin on a system
<kaustuv>
Besides, I cannot distribute cygwin because of license issues. I need to add code to the installer to download it.
<Camarade_Tux>
hmm, right and right =/
<kaustuv>
WiX doesn't seem so bad really. Have you read the tutorial?
<Camarade_Tux>
I only went through it pretty quickly and it looked ok but I really never tried it
vindimy has quit [Read error: 54 (Connection reset by peer)]
<kaustuv>
Camarade_Tux: which POSIX environment on Windows did you say you preferred? ISTR it wasn't Cygwin...
<Camarade_Tux>
kaustuv: well, there's msys but msys is an unofficial and never merged back target for gcc-2.9x (it's a fork of cygwin too), it has nice things but it's also completely annoying to instal
<Camarade_Tux>
l
<Camarade_Tux>
then you have SFU/SUA (services for unix) which works pretty well but is proprietary (derives from bsd tools), it uses a windows service to do what cygwin does with cygwin1.dll
<Camarade_Tux>
so, no clear answer unfortunately
<kaustuv>
Hmm... I really only need Cygwin for Isabelle because my OCaml code compiles with MS compilers fine and doesn't use any Unix tools.
<Camarade_Tux>
what I usually do is native apps running along msys ones (with the shell and a few others), and that means some path separators issues (/ vs \\)
<Camarade_Tux>
isabelle doesn't work as a native app?
<kaustuv>
No, it requires polyml which doesn't compile with mingw/msys
<Camarade_Tux>
and it has a graphical tool, right?
<kaustuv>
Probably, but I only need the Isabelle kernel as I use it as a proof checker
<Camarade_Tux>
(there is xming but everything adds up and it may be really hard to install in the end)
<Camarade_Tux>
kaustuv: what is the problem with polyml?
<kaustuv>
It has mysterious compilation errors with C++ using the new mingw gcc4 release
<kaustuv>
I tried to diagnose the issue, but had to stop because I was losing my sanity
<kaustuv>
I guess it compiles on the gcc3 version of mingw, but then it doesn't work with activestate perl
<Camarade_Tux>
do you still have them?
<kaustuv>
actually, no, I just nuked that VM. But I'll try again using this new Win7 VM that I'm building right now.
<Camarade_Tux>
you may try using mingw-w64 instead of mingw (it's just a matter of extracting the toolchain)
<Camarade_Tux>
kaustuv: is you windows 64bit or 32bit?
<kaustuv>
64
<kaustuv>
but I believe I had the problems earlier with 32bit Vista
<Camarade_Tux>
and is poly/ml written in C/C++ only?
<Camarade_Tux>
kaustuv: mingw-w64 also targets win32 :)
<kaustuv>
Polyml is just an interpreter, so most of it is C++. The standard library and some polyml-specific libraries are in SML.
<Camarade_Tux>
ok, I'll try a cross-compilation, at least some of it should work :)
<kaustuv>
the isabelle wrapper scripts are also written in a mixture of bash and perl, though I think I just have to instrument them to figure out what polyml invocation they use...
<kaustuv>
too bad there isn't an #isabelle
<kaustuv>
Wait, mingw-w64 runs under cygwin?
<Camarade_Tux>
yeah, as a cross-compiler
<Camarade_Tux>
cygwin is 32bit-only
<kaustuv>
Ah, interesting. Thanks for pointing me to this.
<Camarade_Tux>
(well, that's not the reason it only cross-compiles but that's good to know)
<Camarade_Tux>
:)
<Camarade_Tux>
mingw-w64 is also much more reactive than mingw.org (even HURD is more reactive than mingw.org actually)
<kaustuv>
Can it cross compile from {32,64} bit Linux?
<Camarade_Tux>
sure, I'm currently downloading a cross-compilation toolchain :)
<Camarade_Tux>
hmmmm.......................... probably because I said I was going to work on that
<Camarade_Tux>
a promise is a bit strong (although the things to change on inria's source are pretty small)
<kaustuv>
So if I get this WPG System64 preview and install it over MSys 1.0 then magic will happen?
Jedai has quit [Read error: 110 (Connection timed out)]
<kaustuv>
(in Win7 64bit)
<Camarade_Tux>
s/will/may/
<Camarade_Tux>
sorry, that's sed parlance from another channel, it may work, it may not work, wpg system64 has some problems too
<kaustuv>
I guess the website isn't clear if I need MSys already or not before unzipping WPG System64. Would you happen to know?
Snark has quit ["Ex-Chat"]
Yoric has joined #ocaml
<Yoric>
hi
tmaeda is now known as tmaedaZ
<Camarade_Tux>
kaustuv: no, I don't know, but you'll probably want msys anyway and it's probably a separate download, also having msys "see" the mingw installation is quite easy to do by hand if there isn't an installer already (for msys)
<Camarade_Tux>
hi Yoric
tmaedaZ is now known as tmaeda
<gildor>
hi Yoric
ygrek has joined #ocaml
<Camarade_Tux>
hmmm, a problem with mingw-w64, for 32bit, it seems some header has a problem (mingw-w64's compiler is usually better than mingw.org's but the headers haven't been tested as much)
Yokitx has quit [Read error: 110 (Connection timed out)]
struktured has joined #ocaml
<kaustuv>
Awesome, polyml compiled under WPG Sys64
<kaustuv>
Now for the real test -- does it work
<Camarade_Tux>
he, nice :)
<Camarade_Tux>
and my error seems to be a polyml error actually
tonyIII__ has joined #ocaml
tonyIII_ has quit [Read error: 104 (Connection reset by peer)]
animist has quit [Read error: 60 (Operation timed out)]
Yokitx has joined #ocaml
struktured has quit [Read error: 110 (Connection timed out)]
munga has quit [Read error: 60 (Operation timed out)]
Vital303 has quit [Read error: 113 (No route to host)]
verte_ is now known as verte
tmaedaZ is now known as tmaeda
verte has quit ["~~~ Crash in JIT!"]
ygrek has quit [Remote closed the connection]
FrihD has joined #ocaml
<FrihD>
salut
ttamttam has quit [Read error: 110 (Connection timed out)]
struktured has joined #ocaml
ikaros has quit [Remote closed the connection]
<Camarade_Tux>
hi FrihD
<FrihD>
juste pour savoir, ça existe des bindings gobject / ocaml ?
<FrihD>
(et si oui, si c'est crade ou super crade)
<Camarade_Tux>
lablgtk ?
<Camarade_Tux>
what do you need? / de quoi est-ce que tu as besoin plus précisement ?
<kaustuv>
Camarade_Tux: it seems that vanilla ocaml will not compile on WPG Sys64. Did you have some patches for it?
<Camarade_Tux>
kaustuv: iirc the first problem is that it looks for gcc instead of i686-w64-mingw32-gcc or x86_64-w64-mingw32-gcc (and same for as, ld, and all their friends), these are defined in only a few files so you can probably do that by hand
<Camarade_Tux>
after that, I ran into a naming problem: ocaml wanted to use stdlibstdlib.cma (or .a) and I didn't solved that
* Camarade_Tux
checks if he still has his modified files for the first proble
<Camarade_Tux>
m
<kaustuv>
Hmm... I tried what I could find (such as running configure with -host x86_64-w64-mingw32), but then it said something bizarre -- that the architecture was neither big endian nor little endian
<thelema>
Camarade_Tux: you're patching ocaml to use 64-bit mingwin compiler?
<FrihD>
Camarade_Tux: j'étudie si utiliser l'interface gupnp est possible :)
<FrihD>
mais si faut charger tout gtk, nan merci :)
<Camarade_Tux>
kaustuv: you shouldn't use configure on msys/msvc, you should copy a few files (config/{m,s}-nt.h and another one which name I've forgotten) and use 'make -f Makefile.nt'
<Camarade_Tux>
thelema: well, it basically works afaik, it's more a configuration issue
<Camarade_Tux>
FrihD: I don't think gupnp is in lablgtk anyway / je pense que gupnp n'est pas dispo dans lablgtk de toute manière
<kaustuv>
If I want to distribute binaries compiled with mingw32 or msvc, do I have to distribute flexdll also?
andre has joined #ocaml
_andre has quit [Read error: 60 (Operation timed out)]
h3r3tic has quit [Read error: 104 (Connection reset by peer)]
<Camarade_Tux>
FrihD: it may be possible to use only the glib part of lablgtk but I'm not sure / y'a peut-être moyen de n'utilise que la partie glib de lablgtk mais je suis pas sûr
h3r3tic has joined #ocaml
<kaustuv>
offtopic, but is #ocaml-fr dead?
<Camarade_Tux>
FrihD: if you manage to, tell me, I'll maybe be able to provide you with bindings or at least the beginning of them / si tu y arrives, dis-le moi, je peux peut-être te passer des bindings ou un début de
<Camarade_Tux>
kaustuv: dunno, I guess it's less visible (I don't know where the name of the channel is visible)
<FrihD>
Camarade_Tux: thanks, but i don't need it very soon ;) i'm a newcomer to ocaml, it's for a project where I might have to use it
<Camarade_Tux>
kaustuv: about flexdll, no idea, I guess you'll have to test and see by yourself ;-)
<kaustuv>
Camarade_Tux: Since polyml doesn't work in mingw-w64 anyhow (core dumps trying to load saved state), I am quickly losing any motivation to provide Windows native binaries for anything else.
<Camarade_Tux>
FrihD: I'm currently on a project that generates bindings to glib2-based libraries automatically, it's just not complete yet but should be soon, if you need gupnp bindings but these aren't available yet, tell me and I'll check gupnp a bit more
<flux>
kaustuv, how come this sudden interest in polyml?
<kaustuv>
flux: I have a system that produces proofs for Isabelle to check. Isabelle runs on polyml.
<kaustuv>
flux: but everything that produces the proofs is in ocaml, so it's really unfortunate that Isabelle doesn't run on anything else
<flux>
ok
<flux>
port isabelle?-)
<Camarade_Tux>
kaustuv: it seems wpg-system64 uses gcc-4.5 which is probably pretty unstable right now, also poly/ml shouldn't even compile for w64 (from what I've seen)
<kaustuv>
flux: I think that would be a 10 year project and my post doc position expires at the end of this month :(
<kaustuv>
Camarade_Tux: I got polyml to compile but it required some nontrivial code editing that I am quite sure were not entirely sound
<FrihD>
ok, thanks Camarade_Tux
<flux>
quick, someone come up with a camlp4-frontend for sml.. ;)
<kaustuv>
Camarade_Tux: By the way, polyml compiles fine with gcc 4.5 on *nix, so I think the issues are elsewhere
<Camarade_Tux>
kaustuv: with 4.*5*?
<Camarade_Tux>
I mean, it's possible (and maybe even likely) but I'm surprised you have that version on your computer ;-)
<Camarade_Tux>
flux: hehe :P
<kaustuv>
err, no, 4.3.3. I assumed 4.5 was the latest release, but clearly I was wrong
<Camarade_Tux>
4.5 is the release that isn't released yet ;-)
<Camarade_Tux>
hmmm, I'm pondering working on linux with windows inside a virtual machine, making changes and using git from linux but compiling from windows
<kaustuv>
Oh well, I think expecting users who use theorem provers not to be able to run Cygwin is an unrealistic assumption anyway, and to hell with the lawyers. I'm just going to distribute cygwin.dll unmodified as the licence clearly says I can.
<Camarade_Tux>
kaustuv: I still have the source which let me compile most of ocaml, do you want them? (or rather a tarball of the modified files)
<kaustuv>
Camarade_Tux: I can forward them to someone in Gallium if you want, but personally I'm going to stop wasting time on mingw for now.
<Camarade_Tux>
kaustuv: these are pretty trivial changes (but finding which files to change has been a bit harder), but for the remaining part, I have yet to spend more time on it (when I reached that point, I had to stop in order to do other things)
<Camarade_Tux>
I should use windows more in a few days, after ocaml-gir gets in a better state (I have to support out-parameters, which I didn't want to do)
<kaustuv>
I hope that ocamlforge project to have a decent native ocaml compilers in windows produces some useful results. Meanwhile, I am looking forward to never having to use Windows professionally again.
Smerdyakov has joined #ocaml
<Camarade_Tux>
well, windows ain't that annoying by itself, it's the complete lack of package manager
<Smerdyakov>
Does anyone have suggestions of good demos to showcase the usefulness of type-level programming for web applications?
<Camarade_Tux>
I'm using slackware, I don't ask for a package manager that babysits me (and I wouldn't want one), but besides cygwin, there is absolutely nothing and I find myself doing the same things dozens of times =/
<kaustuv>
Smerdyakov: XML normalization by evaluation!
* kaustuv
ducks
<Smerdyakov>
I'm only looking to use XML as far as is needed to get browsers to display pages. :)
seanmcl has joined #ocaml
<kaustuv>
Camarade_Tux: Cygwin is not too bad really, especially since setup.exe -q -P foo is a reasonable approximation for apt-get install foo
romanoffi has quit [Read error: 148 (No route to host)]
<kaustuv>
Camarade_Tux: but I think Windows with a working POSIX environment, a GNU toolchain, and maybe a window manager that supports multiple desktops would almost be usable
<kaustuv>
Clearly Windows has a long way to go before it can succeed on the desktop
<Camarade_Tux>
kaustuv: yeah, windows isn't bad and you can get pretty small environments, but it's slow, you could read, parse and execute two configure.in in your head while a single ./configure is running :P
<Camarade_Tux>
kaustuv: definitely, windows isn't there yet ;p
<kaustuv>
configure is kind of a worst case for cygwin though because it forks like crazy
<Camarade_Tux>
yeah, hopefully cygwin has packages and doesn't let you build everything with configure and make :P
seanmcl has quit []
Smerdyakov has quit ["Leaving"]
valross has quit [Read error: 110 (Connection timed out)]
valross has joined #ocaml
tmaeda has quit [Read error: 60 (Operation timed out)]
ski_ has quit ["Lost terminal"]
tmaedaZ has joined #ocaml
tmaedaZ is now known as tmaeda
ski_ has joined #ocaml
seanmcl has joined #ocaml
seanmcl has quit [Client Quit]
Vital303 has joined #ocaml
antegallya has joined #ocaml
hcarty has quit ["Lost terminal"]
antegallya has quit [Read error: 104 (Connection reset by peer)]
antegallya has joined #ocaml
animist has quit [Remote closed the connection]
animist has joined #ocaml
antegallya has quit [Success]
antegallya has joined #ocaml
antegallya has left #ocaml []
ikaros has joined #ocaml
_zack has quit ["Leaving."]
Associat0r has joined #ocaml
Nutssh has quit ["Client exiting"]
Alpounet has quit [Read error: 113 (No route to host)]
ikaros_ has joined #ocaml
ikaros has quit [Read error: 60 (Operation timed out)]
seanmcl has joined #ocaml
spicey has quit [Read error: 104 (Connection reset by peer)]
<Camarade_Tux>
he, and that reminds me of one reason I don't like cygwin: packages older than me, I think gtk is five years old (and ocaml is 3.08)
seanmcl has quit []
seanmcl has joined #ocaml
<kaustuv>
They're supposed to be releasing 1.7.1 soon, but I doubt with a significantly better situation
<Camarade_Tux>
I think I stopped using cygwin more than three years ago partly because of that, I doubt it'll improve
_zack has joined #ocaml
eydaimon has quit ["WeeChat 0.3.0"]
Ched has joined #ocaml
tarbo2 has quit [Read error: 110 (Connection timed out)]
kaustuv_ has joined #ocaml
rwmjones_ has joined #ocaml
jamesmkukla has joined #ocaml
seanmcl has quit []
FrihD has quit ["Leaving."]
kaustuv has quit [Read error: 110 (Connection timed out)]
ygrek has joined #ocaml
rwmjones has quit [Read error: 113 (No route to host)]
rwmjones_ is now known as rwmjones
seanmcl has joined #ocaml
seanmcl has quit [Client Quit]
valross has quit [Remote closed the connection]
Submarine has joined #ocaml
seanmcl has joined #ocaml
seanmcl has quit [Client Quit]
<Camarade_Tux>
he, even python guys have troubles with gobject-introspection :P
ikaros has quit ["Leave the magic to Houdini"]
ikaros has joined #ocaml
Ched has quit [Remote closed the connection]
seanmcl has joined #ocaml
Ched has joined #ocaml
hcarty_phone has joined #ocaml
ttamttam has joined #ocaml
Yoric has quit [Remote closed the connection]
Snark has joined #ocaml
lutter1 has quit ["Leaving."]
tmaeda is now known as tmaedaZ
ulfdoz has joined #ocaml
jamesmkukla has left #ocaml []
albacker has joined #ocaml
ttamttam has quit ["Leaving."]
munga has joined #ocaml
ulfdoz_ has joined #ocaml
ulfdoz has quit [Read error: 110 (Connection timed out)]
ulfdoz_ is now known as ulfdoz
h3r3tic has quit [Read error: 104 (Connection reset by peer)]
h3r3tic has joined #ocaml
h3r3tic has quit [Read error: 104 (Connection reset by peer)]
h3r3tic has joined #ocaml
smimou has joined #ocaml
hcarty has joined #ocaml
onigiri has joined #ocaml
ulfdoz has quit [Read error: 60 (Operation timed out)]
Yoric[DT] has joined #ocaml
Submarine has quit [Read error: 145 (Connection timed out)]
_JusSx_ has joined #ocaml
Submarine has joined #ocaml
Ringo47 has quit [Read error: 60 (Operation timed out)]
tmaedaZ0 has joined #ocaml
smimou has quit [Read error: 110 (Connection timed out)]
ttamttam has joined #ocaml
smimou has joined #ocaml
rbancroft has quit [Client Quit]
tmaedaZ has quit [Read error: 110 (Connection timed out)]
Ringo48 has quit [Read error: 110 (Connection timed out)]
tarbo2 has joined #ocaml
julm_ has joined #ocaml
julm has quit [Read error: 60 (Operation timed out)]
rbancroft has joined #ocaml
Alpounet has joined #ocaml
ttamttam has quit ["Leaving."]
julm_ is now known as julm
andre has quit ["leaving"]
_zack has quit ["Leaving."]
h3r3tic has quit [Read error: 54 (Connection reset by peer)]
h3r3tic has joined #ocaml
ulfdoz has joined #ocaml
Snark has quit ["Ex-Chat"]
Associat0r has quit []
Submarine has quit [Remote closed the connection]
ulfdoz has quit [Read error: 110 (Connection timed out)]
<Camarade_Tux>
on the C side, if I use: [ value uple = caml_alloc(1, 0); Store_field(uple, 0, Val_int(1)); ], what do I get? caml_alloc() is meant to create tuples, but what if I make a 1-uplet with that? could this create problems?
<julm>
Camarade_Tux: I don't think so since it's used in asmrun/natdynlink.c: res = caml_alloc(1,0);
<julm>
note that there also is caml_alloc_small available
* Camarade_Tux
is relieve he won't have to write spaghetti code
seanmcl has quit []
<Camarade_Tux>
*relieved
seanmcl has joined #ocaml
hcarty_phone has quit [Read error: 110 (Connection timed out)]
_JusSx_ has quit ["leaving"]
smimou has quit [Read error: 110 (Connection timed out)]
ygrek has quit [Remote closed the connection]
<Camarade_Tux>
oh, forgot that, turns out a 1-element caml_alloc is most probably a record (which means I'll have to spaghetti-code)
<julm>
*miam*
<Camarade_Tux>
good idea for tomorrow :)
<Camarade_Tux>
well, it could also be an array rather than a record
<Camarade_Tux>
uniform representation is nice until you have to list all the things that could look alike :P
smimou has joined #ocaml
ikaros has quit ["Leave the magic to Houdini"]
ikaros has joined #ocaml
albacker has quit ["Leaving"]
Yokitx has joined #ocaml
h3r3tic has quit [Read error: 104 (Connection reset by peer)]