sponge45 changed the topic of #ocaml to: Discussions about the OCaml programming language | http://caml.inria.fr/
Axioplase has quit ["leaving"]
Skal has quit [Read error: 104 (Connection reset by peer)]
pango_ has joined #ocaml
mbishop has joined #ocaml
pango has quit [Remote closed the connection]
<Shimei> So, I hear a lot about the "revised syntax" and I'm reading the chapter on it (http://caml.inria.fr/pub/docs/manual-camlp4/manual007.html). Is it syntax slated for a future release, or is there some option to enable it?
<Shimei> Oh, nevermind. I figured it out.
<Smerdyakov> I don't think the revised syntax has any serious traction outside the mind of the campl4 author, and just serves as a nuisance. ;)
Demitar_ has joined #ocaml
<Shimei> I like some of it, but the pattern matching syntax seems particularly nasty (the brackets).
<Smerdyakov> Liking it or not is not very relevant. There is too much OCaml code already out there for it to be practical to switch.
<Shimei> Yeah. I imagine so. Couldn't it be gradually phased in if it's introduced while maintaining backwards compatibility though?
joshcryer has quit [Read error: 131 (Connection reset by peer)]
<Smerdyakov> I think that's almost 100% unlikely given the social factors.
joshcryer has joined #ocaml
Demitar has quit [Read error: 60 (Operation timed out)]
Eridius has quit [Connection timed out]
ikaros has quit [Read error: 110 (Connection timed out)]
ikaros has joined #ocaml
sponge45 has joined #ocaml
whatsup103 has joined #ocaml
danly has quit [Read error: 104 (Connection reset by peer)]
<whatsup103> back to trying to make my csv stuff
mbishop has quit [Remote closed the connection]
Shimei has quit [Read error: 110 (Connection timed out)]
Demitar_ has quit [Read error: 104 (Connection reset by peer)]
Demitar__ has joined #ocaml
Demitar__ has quit [Read error: 104 (Connection reset by peer)]
Submarine has quit [Remote closed the connection]
<whatsup103> i have this csv library .........where do i place it ? usually every library i install .......when i do make install / my package system it puts it in the right location
<whatsup103> but yea i donmt know where ocaml libs are suppose to go
<stevan_> whatsup103: I installed the OS X binary from http://caml.inria.fr/ and all the libs are in /usr/local/lib/ocaml/stie-lib
<stevan_> your milage may vary
<whatsup103> dark_light: that csv thing provided many examples including ways how they iter'd things
dibblego has left #ocaml []
romanoffi has joined #ocaml
twobitsprite has quit [Read error: 110 (Connection timed out)]
gunark has quit [Nick collision from services.]
gunark has joined #ocaml
gunark has quit [Nick collision from services.]
<whatsup103> there was some FLTK bidnings floating around somewhere anyone know where they are ?
mbishop has joined #ocaml
<mbishop> is it just me, or is the "revised syntax" a lot like SML?
<whatsup103> im stil stumped how to dump this into the DB now hmmmmmm
<whatsup103> i got it to read and iter just need to dump now
microcape has quit []
ChoJin has joined #ocaml
sponge45 has quit ["zzzzzzzzzz"]
romanoffi has left #ocaml []
piggybox has quit [Connection timed out]
piggybox has joined #ocaml
<dark_light> whatsup103, :)
ChoJin has quit ["This computer has gone to sleep"]
* whatsup103 installs MySQL-Ocaml bindings
<whatsup103> dark_light: the mullata women on the beaches are beautifuuuul
<dark_light> aahahahaha oh man
<dark_light> they are beautiful, but this effect on tourists is very bad for brazil :-(
<dark_light> but, whatever.. :P
<whatsup103> what?
<whatsup103> dark_light: mullata is different from mullata in usa
<whatsup103> mullata in usa means like pretty much black with some white
<whatsup103> here i see tan ass white girls calling themselves mullata
<dark_light> well i don't use really the term mullata
<whatsup103> dark_light: 2ndly im no tourist :D ..........i live in rio .......but i goto some tourist spots every now and then
<dark_light> but the term here means black+white too :P
<whatsup103> mullata in usa is highly racist to say lol
<mbishop> Really?
<mbishop> never heard of it
<whatsup103> yes
<whatsup103> very racist
<whatsup103> mbishop: where do you live ?
<mbishop> Louisiana, so you'd think I'd know all about racist terms :P
<whatsup103> what part of lousi,
<mbishop> actually, mullata is starting to sound kind of familiar, but never heard anyone say it
<mbishop> the south...about 60 miles south of new orleans
<whatsup103> i see
<whatsup103> I prefer Miami
<whatsup103> :D
<whatsup103> dark_light: also i think it was you that told me rio de janeiro was all black and no white peole was it ?
<whatsup103> dark_light: half the city is white
<dark_light> well i don't said that
<dark_light> but i don't know exactly what is white
<whatsup103> lol
<dark_light> actually for me white is RGB #FFFFFF and i didn't know anyone that is white. lol
<whatsup103> lol
<dark_light> i really prefer do not refer skin color óo
<whatsup103> maybe it wasnt you ;0
<whatsup103> im peach :D
<whatsup103> netter :)
<whatsup103> better
<dark_light> whatsup103, here racism is a bit hidden, too :P
<dark_light> i think i am yellow
<dark_light> i look like yellow
<dark_light> :P
<whatsup103> lol
<whatsup103> asian :D
<whatsup103> my origianl skin color is like peach .........but being in rio will make you bronze
<dark_light> i really hate the sun, if i liked it i would become a bit black i think
<whatsup103> move to new york city
<whatsup103> you will become light i promise
<dark_light> yeah :(
<whatsup103> might take a few years :D
<whatsup103> casue i am a very light peach color normally ..............not albino or sickly looking ;) but i have been living in miami and puerto rico for years so i am very dark ...........and i moved ot canad for 2 years and still never got my original skin color back
<whatsup103> but i like what i got now so yea :D
<whatsup103> im suprised i see tanning salons in rio
<whatsup103> ........thats the stupidest thing ever
<dark_light> tanning salons?
<whatsup103> thing girls goto and lay in a bed and get dark
<dark_light> ahahahaha
<whatsup103> you know ?
<dark_light> yeah
<whatsup103> rio is covered with them lol
<dark_light> it is strange, but.. :o
<whatsup103> lol
<whatsup103> the sun is free :D
<dark_light> if rio was very cold, i could understand it
<dark_light> yeah :P
<whatsup103> me too
<whatsup103> same way in miami though miami is almost like rio weather wise
<whatsup103> all them cuban girls goto the tanning salons like crazy
<dark_light> but, well, it's more likely i want get dark then i want go to the beach :P
<whatsup103> i goto beach 2-3 days a week
<whatsup103> play volleyball
<dark_light> (but well maybe someday i will change my mind about beachs..)
<whatsup103> you dont goto beach ?
<dark_light> never
<whatsup103> lol
<dark_light> lol
<whatsup103> damn
<whatsup103> beach is fun ........i play volleyball with alot of women :D
<dark_light> it's very hot. :(.
<whatsup103> and when i first got there i didnt speak porturgese all to well .........and they all wanted ot talk ot me and some of the guys started calling me gringo n shit
<whatsup103> but this girl across the street i have to help her with her english for school :D
<whatsup103> :X
<whatsup103> shes real hot
<whatsup103> although the name Rafaela ...........i dont like to well ...........reminds me of a ninja turtle
<dark_light> ahahahhaha
<whatsup103> you remember ninja turtles ?
<dark_light> yeah
<dark_light> donatello
<dark_light> well, almost
<whatsup103> and Rafael
<dark_light> but rafaela is a fine name for me
<whatsup103> yea i like it now too but still :D
<whatsup103> it reminds me of ninja turtle :D
<whatsup103> out of all the places i have lived the usa and brazil have more in common then any other country i have saw
<dark_light> whatsup103, but still it's strange for me if a brazillian, in brazil, wants to talk with a american in english. :P
<dark_light> sorry :P
<whatsup103> what you mean ?
<whatsup103> its just curiocity i guessss
<whatsup103> alot of people want to talk to me in english
<whatsup103> at store and stuff
<dark_light> whatsup103, i see it like.. lack of nacionality
<dark_light> but, whatever :P
<dark_light> nacionalism
<whatsup103> nationality ?
<dark_light> love for it's own country
<whatsup103> ummmmmmm
<whatsup103> lol
<whatsup103> language has nothing to do with the country
<dark_light> love for it's own language
<dark_light> lol
<whatsup103> people want to know more then one :D
<whatsup103> you know english well ..........whats wrong with that :D
<dark_light> i am in a international channel, it's different :P
<dark_light> :PPP
<whatsup103> lol
<whatsup103> soo you dont like sun no beach .........
<whatsup103> hate other languages
<whatsup103> :( lol
<dark_light> you are exaggerating :P
<whatsup103> :D
<whatsup103> didnt you say you wanted ot move to switherland one time tooo ?
<whatsup103> lol
<dark_light> yeah, and there i would speak whatever the language they speaks
<whatsup103> lol \
<dark_light> that's what i am saying
<whatsup103> << hate switcherland
<whatsup103> socilist paradise
<dark_light> i don't know really what country is this
<dark_light> ahn..
<whatsup103> lol
<whatsup103> its a socilist paradise government takes all your money
<dark_light> it's in europe? :P
<whatsup103> yes
<whatsup103> lol
<whatsup103> i am business owner
<whatsup103> i dont like government taking 70 percent of my money
<dark_light> ok, i really don't know what you are talking about
<dark_light> hmm
<whatsup103> the country lol
<whatsup103> lol
<dark_light> oh i don't know much of swiss
Smerdyakov has quit ["Leaving"]
<whatsup103> oo
<whatsup103> look at map of europe and how small all of it is
<dark_light> they speaks german, italian, french and romanish? nice=)
<dark_light> yeah
* dark_light likes small nations
<whatsup103> the whole european union could fit into the usa
<whatsup103> maybe brazil lol
<dark_light> whatsup103, that's exactly why they can manage themselves well
<dark_light> brazil is a very corrupt nation
<whatsup103> i dis agree
<dark_light> whatsup103, what about talking in PVT? lol :P
<whatsup103> depemnds on where you live
<whatsup103> :D
_velco has joined #ocaml
buluca has left #ocaml []
khaladan has joined #ocaml
khaladan_ has joined #ocaml
love-pingoo has quit ["Connection reset by pear"]
ChoJin has joined #ocaml
khaladan has quit [Connection timed out]
khaladan_ is now known as khaladan
slipstream has joined #ocaml
johnnowak has joined #ocaml
slipstream-- has quit [Read error: 110 (Connection timed out)]
shawn has quit ["This computer has gone to sleep"]
khaladan has quit [Connection timed out]
pango_ has quit [Remote closed the connection]
velco has joined #ocaml
pango has joined #ocaml
ikaros has quit [Read error: 110 (Connection timed out)]
johnnowak has quit []
shawn has joined #ocaml
pango_ has joined #ocaml
pango has quit [Remote closed the connection]
piggybox has quit []
shawn has quit [Operation timed out]
love-pingoo has joined #ocaml
shawn has joined #ocaml
<flux__> hmm.. the ocaml compiler (and why not other compilers too) could make educated guesses about the reasons of the error based on the indentation of the code..
<flux__> like if the same line is indented to begin at the same column as the previous line, and the previous line misses a ;, and the previous line also has a function-specific type error, it could suggest that maybe ; is missing
joshcryer has quit [Excess Flood]
joshcryer has joined #ocaml
ramkrsna has quit [Read error: 54 (Connection reset by peer)]
<love-pingoo> flux__: I guess that would be a huge change, since the lexer/parser completely forgets about whitespace
<flux__> hm, good point
<flux__> but would it be a huge change to store it somewhere?
<flux__> although the code does get mangled quite a bit
<flux__> hey, all pieces of code have their locations, right?
<love-pingoo> that's right
<flux__> so it could simply scan the nearby elements and look for their locations
<love-pingoo> so the indentation could be retrieved from that
ppsmimou has joined #ocaml
ChoJin has quit ["This computer has gone to sleep"]
_fab has joined #ocaml
Submarine has joined #ocaml
Skal has joined #ocaml
<whatsup103> whats that one website where alot of ocaml projkects are stored
<whatsup103> im looking for the ocaml-FLTK bidning
<whatsup103> binding
<flux__> I always google for 'caml hump'
<whatsup103> not that
<whatsup103> theres another one
<whatsup103> kinda like a sourceforge
<flux__> yes, I recall seeing another one too..
<flux__> please find and tell :-)
<whatsup103> looking for that ocaml fltk thing
malc_ has joined #ocaml
<whatsup103> flux__: do those binding generators work well ?
<flux__> I haven't really used any
<whatsup103> casue they are using them for D for certain things
<whatsup103> so i wonder if it would wor kas well for Ocaml
<whatsup103> be cool to use QT or whatever you want in it
<whatsup103> i've been seeing this Felix programming language popping up everywhere
<whatsup103> like some kind of ML to c++ or something
<whatsup103> its getti quite a bit of support lol
<whatsup103> NEW Fast Asynchronous I/O support provides transparent pre-emptive threads, timers and
<whatsup103> sockets optimised for your platform. Using epoll on Linux, kqueue on OSX and BSD,
<whatsup103> and IO completion ports on Windows. Your networks apps never ran faster!
<whatsup103> my im client envies that
<pango_> you don't need anything like that in an im client
<whatsup103> The Felix compiler is written in Objective Caml, and generates ISO C++ which should compile on any platform.
<pango_> unless you continuously chat with 10k+ people simultaneously
<whatsup103> pango_: but you know it would still be coool
<pango_> coolness leave me cold
<whatsup103> what you guys think of that
<whatsup103> ^^
<pango_> the only mention of ocaml-fltk bindings on the web is in an article that uses it as an example of weird library dependancy... Seems it was invented for the argument, there's no sign of such thing already existing
<whatsup103> hmmmmmm
<pango_> it's used to teach compilation techniques
<whatsup103> isee
<whatsup103> so could it be better then the regualr compiler ?
<pango_> it probably has 10% of the regular compiler features, so yes, if you only use those 10%...
<whatsup103> :D
<whatsup103> that looks real tight
<whatsup103> if it works
<pango_> C is not an excellent lowlevel language for compilers, that's why ocamlopt generates asm and GHC uses C--
<pango_> (at least not for all languages)
<pango_> oh, my bad, that's yet another bindings generator
<pango_> got confused by the name ;)
<whatsup103> lol
<whatsup103> if the bidning generator works and doesnt add any massive size i would use it
<pango_> no update since the initial checkin 5 years ago... not unusual in Sourceforge, but still...
<pango_> better use something else
<malc_> 2006 - 2000 = 6
<whatsup103> if it works it works
<whatsup103> just a binding gernator ........
<pango_> viewCVS doesn't know how to count then ;) http://c2caml.cvs.sourceforge.net/c2caml/c2caml/
<pango_> whatsup103: there's no such thing as "just a" bindings generator
<whatsup103> ?
<malc_> i guess someone should write something based on `-fdump-translation-unit'
<whatsup103> thanks for the link
<whatsup103> epoll is built into the linux kernnel oooooo
<flux__> malc_, how would -fdump-translation-unit help?
<flux__> atleast it doesn't generate anything on struct definitions
<malc_> drag some variable with that struct type in scope and it will
<malc_> anyhow
malc_ has quit ["leaving"]
<whatsup103> w0000t
<whatsup103> libevent
<whatsup103> :D
<whatsup103> epoll
<pango_> whatsup103: painting is still fresh ;)
<whatsup103> lol
<whatsup103> what?
<whatsup103> since only 2.6+ comps use epoll anyways prolly wouldnt be smart to even use it
<flux__> but the library provides backward compatibility, doesn't it?
<flux__> also I imagine these days 2.6 is quite common..
<flux__> atleast if you want performance
<whatsup103> its only 1 connection
<whatsup103> an im client
<whatsup103> :D
<flux__> well, the interface might still be convenient
<whatsup103> what do you mean ?
<whatsup103> i mean theres a libevent that provides it just fine
<whatsup103> i was trying to find a gtk+ patch that provides it
<pango_> whatsup103: I already mentionned libevent to you (as youknow365 at the time :) ) September 3rd ;)
<whatsup103> ooooooo
<whatsup103> :D
<whatsup103> i must have skiffed over it
<pango_> but even at the time it was overkill for an im client ;)
<whatsup103> well i got it done now with select
<whatsup103> ./ poll
<whatsup103> but i could always improve it
<flux__> hmm.. about my 'how to figure out current function in camlp4'-problem, would the newsgroup or the mailing list suit it better?
<whatsup103> i actually use the gtk library to do it
<whatsup103> but heh
<whatsup103> your right using epoll would be an overkill and i would loose the multiplatformnesss
<flux__> I actually want to make a macro that checks if the name of the current function may be found from a hashtable, and if so, perform an operation.. like DEBUG "hello, %s" msg
<flux__> although I suppose in the first stage I could just forget the hash table
<pango_> whatsup103: not thru libevent, that uses what's available on current platform
<whatsup103> i wanted ot use
<whatsup103> the gtk+ way though
<whatsup103> libeven was a good find
<whatsup103> but would add more file size to the app and another required lib to include
<whatsup103> probably wouldnt add to much but the benefit doesnt outweigh
DRMacIver has quit [Read error: 104 (Connection reset by peer)]
<whatsup103> theres a bug open now to add epoll support so im guessing it probably will get released in there in a little while anywho so yea
DRMacIver has joined #ocaml
romanoff1 has joined #ocaml
Submarine has quit [Remote closed the connection]
holdenk has quit [Remote closed the connection]
terpstra has joined #ocaml
smimou has joined #ocaml
klapmuetz has quit [Read error: 60 (Operation timed out)]
love-pingoo has quit ["Leaving"]
buluca has joined #ocaml
ChoJin has joined #ocaml
velco has quit ["Ex-Chat"]
llama32 has joined #ocaml
<llama32> are any of the ocaml JIT projects actually working on current releases and in continued development?
<haelix> llama32: didn't know there were several of them
<haelix> I only knew about ocamljit. Do you have links to any others ?
<llama32> haelix: i thought i'd run into a few dead[ish?] projects, but it was probably just one
<haelix> :)
<haelix> well as usual you're free... to pick one up and make it livish again.
Smerdyakov has joined #ocaml
<haelix> Or, you may stick with F#
<Smerdyakov> Worse fates exist!
<llama32> does F# work on mono?
<haelix> no idea. I'd be surprised that assembly zouldnot work, though
<romanoff1> according to comments at http://blogs.msdn.com/dsyme/archive/2006/11/30/f-1-1-13-now-available.aspx, F# does work on mono
<romanoff1> have not tried it myself tho
<haelix> Am I the only one dreaming here that closures,
<haelix> In a JITted environnment
<llama32> im worried that since it's got closures and such, it might be more dependant on microsoft's implementation of .NET - whereas mono might not have such things working in a stable state
<haelix> might be partially evaluated
danly has joined #ocaml
<haelix> and be further optimized
<llama32> hmm
<haelix> in order to squeeze a bit more performance
<haelix> think of a runtime constructed (which profiling shows is heavily used) DFA,
<haelix> btree,
<haelix> whatever
<llama32> F# looks interesting [i was just looking at MLj and SML.NET which are similar i guess]
<terpstra> the MS research website for F# is down.. lol
ChoJin has quit ["This computer has gone to sleep"]
<llama32> i think i'd prefer using SML.NET though - simply because it's not from microsoft [haven't seen F#'s license, but i assume it's closed source, and SML.NET's looks pretty liberal]
<llama32> heh, must be running windows
<llama32> does MetaOCaml have anything to do with JIT? i don't really know what multi-stage compilation is...
<terpstra> sml.net is from MS research ...
<llama32> it is?
<terpstra> and only works on windows
<terpstra> yes
<llama32> i can't find that written anywhere
<llama32> not MLj too i hope
<llama32> ugh
<terpstra> "SML.NET is a Standard ML compiler developed at Microsoft Research Cambridge"
<terpstra> MLj is not maintained afaik
<Smerdyakov> Modulo the quite unfortunate problem of being closed source, you ought to take the fact that something is developed at MSR as a sign of high quality, not something to complain about.
<llama32> hmm, how come it has a microsoft.com and a non-microsoft.com website?
<terpstra> MS research works with other universities
<terpstra> some of the guys who work there are pretty cool
<terpstra> they even do P2P research ;-)
<terpstra> it's a fact of life that the only big business sponsored research group is MS research
<terpstra> sad, but true
<llama32> hmm, is F# available as source too?
<llama32> with a "do as you will, but don't blame me" license
<terpstra> i assume it builds on ocaml code, so it presumably has the same license
<haelix> terpstra: you're a bit tough on that. IBM and HP do fund their bit of research
<haelix> IBM: alphaworks
<terpstra> haelix, both of them have been cutting funding
<terpstra> whereas MS has been increasing it
<haelix> HP Labs
<terpstra> look -- i am no fan of MS myself
* llama32 hates microsoft :)
<llama32> and firefox :| which just froze :|
<terpstra> one can hate MS, and still appreciate the guys at MS research tho
<haelix> I prefer Linux, have no special feeling about windows, but MS Research sure looks like a sexy place to work at
<terpstra> often they even GPL their work
<terpstra> yeah, i've thought about it oo
<terpstra> but then i'd have to explain to all my friends that i hadn't REALLY gone to the dark side
<terpstra> i can't find anything about the license on http://strangelights.com/fsharp/Wiki/default.aspx
<llama32> google seems to put some effort into research too
<Smerdyakov> terpstra, I find that _almost_everyone_ who works at MSR is cool. Do you have data to the contrary? :)
<terpstra> nope
Shimei has joined #ocaml
<Smerdyakov> terpstra, which MSR projects are you thinking of that went open source? I know they have a length legal process as a prerequisite for doing that.
<Smerdyakov> s/length/lengthy
<terpstra> i didn't realize google had a research team?
<Smerdyakov> Google has a tiny pool of employees designated as researchers.
<terpstra> i thought they just hired people with PhDs and gave them one day a week to do their own thibng?
<llama32> is there any real difference between SML [+ SML.NET OO extensions] and OCaml [F#]? other than just differently named functions and such?
<Smerdyakov> They make offers to blindingly smart people with PhDs without giving them the chance to enter that set of researchers.
<llama32> terpstra: yeah, i think they give 20% free research time or something
<Smerdyakov> llama32, yes, huge differences, but I don't have time to enumerate them now.
<terpstra> Smerdyakov, i didn't know
<terpstra> sml.net extends sml in many ways
<terpstra> which is why i don't much like it; sml is nice b/c it's simple
<Smerdyakov> terpstra, _do_ you have example of GPL'd software from MSR?
<terpstra> oh, sorry - overlooked that
<Smerdyakov> terpstra, and it has to be software developed when the involved parties were all at MSR.
<terpstra> well, bsd
<haelix> Smerdyakov: I know about the "detour" library
<terpstra> i am not 100% sure about gpl
<terpstra> let me check
<Smerdyakov> terpstra, so anything about GHC can't count. :)
<haelix> which is at least open source
<terpstra> oh
<terpstra> and i don't know how to tell appart "all parties at msr" vs. some??
<haelix> (something to do with win32 call interception. Don't realize how big a deal this is, but this is an open source msr project)
<Smerdyakov> terpstra, I add that particularly to exclude GHC. I think SPJ made a special deal with them to allow him to work on this established open source project.
<terpstra> research.microsoft.com is down :-P
<Smerdyakov> haelix, OK.
<terpstra> i was going to check splitstream and pastry
<terpstra> but -- no website ;-)
<llama32> terpstra: microsoft software running on servers is a fairly ambitious research project afterall
<haelix> Smerdyakov: perhaps more significant is Rotor (the open source .Net framework form MS. Should even compile and run on MacOs)
<llama32> terpstra: can't expect anything to really work until 2040-ish
<terpstra> there was that ms installer too
<Smerdyakov> haelix, I don't think that's MSR.
<terpstra> but i think that's bsd
bluestorm has joined #ocaml
<terpstra> ok, pastry is "BSD-like" not GPL
<haelix> Then... what I'd like to get my hands on are "Bartok" (a fully managed C# OS)
<haelix> and Terminator
<terpstra> so, you could be right about there being none that are gpl
<haelix> (a program terminationo rpoofer)
<terpstra> they may all be bsd
<Smerdyakov> I'm flexible about which open source license is involved. I know that you have to go through Hell and back to release anything open source from MSR today.
<haelix> but well.. not even binaries available :)
<terpstra> it's hard to check via google cache :-P
<Smerdyakov> haelix, Bartok is a compiler, not an OS. I think you mean Singularity.
<terpstra> fix your servers MSR!
<llama32> oddly enough, wikipedia tells me Rotor runs on "Windows, MacOSX and FreeBSD"... i mean FreeBSD is understandably similar to OSX, but seems odd to mention that it works on FreeBSD without saying anything about wether or not it runs on linux
<haelix> terpstra: no they don't - so we can fantasize all we can !
<llama32> perhaps someone got the BSD license confused with FreeBSD
<haelix> Smerdyakov: you must be right
<terpstra> llama32, why do you want an ML running on .NET / JVM?
<Smerdyakov> haelix, I should be; I interned in that project. :)
<haelix> ok :) Congrats for that, btw.
<llama32> terpstra: i sorta like .NET's language-independent-ness, plus there aren't [to my knowledge] any decent/fast/well-supported ml compilers that produce dynamically linkable code with some sort of OO capability, but SML.NET, F# and MLj do this
<llama32> http://www.pps.jussieu.fr/~montela/ocamil/documentation.html <-- anyone know if that is active and/or working?
<Smerdyakov> llama32, is it a port of AbstractIL?
<llama32> no idea
<dark_light> this function http://nopaste.tshw.de/11662060392a6a8/ is tail recursive?
<dark_light> (i suppose isn't)
<Smerdyakov> dark_light, you mean 'read'?
<dark_light> hmm?
<Smerdyakov> dark_light, the function that you are asking about is named 'read'?
<dark_light> ah, yes
<Smerdyakov> dark_light, no, it isn't tail recursive, because of the exception handler.
<dark_light> (but ok the whole function isn't working btw)
<terpstra> cool - F# should work on mono according to the wiki
<Smerdyakov> terpstra, someone already said that here.
<terpstra> didn't see that
<dark_light> Smerdyakov, if i transform this into a type like Value of 'a | Except_EOF, the function could be recursive?
<dark_light> ops.
<dark_light> tail recursive
<Smerdyakov> dark_light, I think I get what you mean, though it's not really clear; you will see a similar thing suggested in some OCaml FAQ.
* llama32 is off to bed
llama32 has quit ["Leaving"]
<dark_light> Smerdyakov, i don't like having to transform exceptions in concrete/variant/etc types
<terpstra> anyways, bye all
terpstra has quit ["Leaving"]
<dark_light> by..
<dark_light> all gone
shawn has quit ["This computer has gone to sleep"]
buluca has quit ["Leaving."]
romanoff1 has left #ocaml []
postalchris has joined #ocaml
shawn has joined #ocaml
shawn has quit [Remote closed the connection]
<mellum> Is there some simple way to implement a purely functional union/find with reasonable performance?
<Smerdyakov> You can always implement any imperative algorithm with only a log factor speed degradation. If that's reasonable to you, then you have your answer.
<dark_light> Smerdyakov, you are serious?
<Smerdyakov> dark_light, yes.
<dark_light> Smerdyakov, i though functional algorithms could have ever at leat the same complexity of imperative equivalents
<Smerdyakov> dark_light, would you mind rewriting that line with attention to proper English grammar? I can't understand it. ;)
<dark_light> Smerdyakov, the problem is: i don't know the proper english grammar, i am a bad student of english at brazil
<dark_light> but say me the parts you didn't understand well :D
<Smerdyakov> dark_light, I see two possible fixings of what you said:
<Smerdyakov> Choice A) "I though functional algorithms could always be as time-efficient as imperative algorithms."
<Smerdyakov> Choice B) "I though some problems intrinsically have much faster imperative solutions than functional solutions."
<Smerdyakov> s/though/thought
<dark_light> Smerdyakov, choice A
<Smerdyakov> dark_light, I don't know how to disprove the statement, but I think it's widely belived to be false.
<dark_light> very nice :)
<dark_light> i mean: many important computer science statments are only beliefs (like NP = P), and that is nice
<dark_light> a young science :)
<bluestorm> hum isn't the common belief than NP != P ?
<mellum> Smerdyakov: Well, I just thought there might be something more elegant.
<dark_light> bluestorm, exactly, common belief
<bluestorm> hm mellum
<bluestorm> i one time read about a "zipper" thing, who looked a bit like the union/find path compression
<bluestorm> but it may be completely wrong as well
<bluestorm> hum
<bluestorm> here
<mellum> bluestorm: Hmm.
<mattam> mellum: you can have a very efficient persistent union/find implemented with side-effects, http://www.lri.fr/~filliatr/software.en.html
<mellum> mattam: Hmm, that can only work on 1..n, it seems... although of course I could add another layer around it
<mattam> what's the problem with that ?
<mellum> mattam: it's not what I need :)
<pango_> maybe implemented as a Set of edges ? (ordered by source node)
<pango_> edge lookup complexity will by O(log n), though
manby-ace has joined #ocaml
<pango_> mmmh probably a Map of edges
manby-ace has left #ocaml []
love-pingoo has joined #ocaml
<pango_> aha. And find must return both the element's set and the "optimized" union-find set
bluestorm is now known as bluestorm_aw
Submarine has joined #ocaml
<pango_> mellum: http://nopaste.tshw.de/11662149921588a/ (not optimized or anything, but you get the idea ;) )
ChoJin has joined #ocaml
Submarine has quit [Remote closed the connection]
romanoffi has joined #ocaml
malc_ has joined #ocaml
gunark has joined #ocaml
Submarine has joined #ocaml
hcarty has joined #ocaml
asm has joined #ocaml
mbishop has quit [Read error: 104 (Connection reset by peer)]
_fab has quit [Read error: 104 (Connection reset by peer)]
stevan_ is now known as stevan
<stevan> I have a really simple question, but I cannot seem to find an answer anywhere
<stevan> I am getting the error "Reference to undefined global `ExtLib'"
<stevan> when I try and use a .cma file I compiled
<Smerdyakov> Did you read the first page of the ExtLib documentation?
<stevan> it was compiled with "ocamlfind ocamlc -package extLib -c *.mli"
<Smerdyakov> I don't know about ExtLib in particular, but every library bundled with OCaml answers this question on its first page of documentation.
<stevan> Smerdyakov: thanks, I will check
<stevan> Smerdyakov: it doesn't say anything regarding this
<stevan> I suspect I am just missing a compiler flag somewhere maybe
<Smerdyakov> stevan, then pick a random library bundled with OCaml and reads its first page of documentation,.
<malc_> stevan: a) what's the point in compiling .mli s only?
<malc_> stevan: b) obviously the error is linker, yet you only show how you _compile_ (interfaces)
<stevan> malc_: sorry, that was an incorrect paste,.. i compile both mli and ml
danly has quit ["Leaving"]
<stevan> malc_: I am using OCamlMakefile, and not doing anything fancy,
<malc_> stevan: basically what the linker has told you, is that it can't find a global named 'ExtLib' in all of the objects libraries that has been supplied
<malc_> hence, extlib.cma (or whatever the name, never used it) is missing
<malc_> correction. is missing or is incorrect position w.r.t. objects that reference said global
<stevan> ah
<stevan> hmm, that might be the difference
<malc_> i'd go and try to make something minimal work.. like file consisting of just open ExtLib (or whatever)
<Smerdyakov> stevan, did you read the first page of documentation of any library bundled with OCaml?
twobitsprite has joined #ocaml
<stevan> Smerdyakov: yes, many,.. none specifically reference my problem nor give any clues
<Smerdyakov> stevan, OK, so lines like this aren't relevant?
<Smerdyakov> Programs that use the unix library must be linked as follows:
<Smerdyakov> ocamlc other options unix.cma other files
<twobitsprite> So... I am completely stumped on my ocamlyacc thinger still... I've worked through ever possible parsing scenario that I can think of and have no idea why I have conflicts...
<twobitsprite> I've also reordered everything every possible way I can think of... left recursion, right recursion, etc
<stevan> Smerdyakov: there are lots of modules bundled with Ocaml, if you had a specific one (and a specific line within one) in mind, it would be have been helpful ifyou mentioned it
<Smerdyakov> stevan, no, I said _libraries_, not modules.
<stevan> Smerdyakov: it seems from the makefile output that extlib is not being inlcuded when the library is linked
<Smerdyakov> stevan, every _library_'s documentation starts with that approximate text.
<stevan> Smerdyakov: thank you
<stevan> I think my issue is more related to OCamlMakefile actually
_velco has quit ["I'm outta here ..."]
Submarine has quit [Read error: 113 (No route to host)]
<stevan> Smerdyakov: I found it, I was missing a flag for OCamlMakefile to pass it into the linker
<stevan> this will teach me to watch my makefile output closer :)
romanoffi has quit [Read error: 110 (Connection timed out)]
whatsup103 has quit [Read error: 104 (Connection reset by peer)]
whatsup103 has joined #ocaml
romanoffi has joined #ocaml
pango_ has quit ["Leaving"]
pango has joined #ocaml
batdog is now known as batdog|gone
bluestorm_aw has quit ["Konversation terminated!"]
twobitsprite has quit [Read error: 110 (Connection timed out)]
smimou has quit ["bli"]
twobitsprite has joined #ocaml