Taaus changed the topic of #ocaml to: http://caml.inria.fr/oreilly-book/
malc has quit [carter.openprojects.net irc.openprojects.net]
gluck has quit [carter.openprojects.net irc.openprojects.net]
malc has joined #ocaml
gluck has joined #ocaml
graydon has joined #ocaml
gluck has quit [Read error: 104 (Connection reset by peer)]
gluck has joined #ocaml
engstad has joined #ocaml
malc has quit ["no reason"]
graydon has quit [Read error: 113 (No route to host)]
graydon has joined #ocaml
engstad has quit ["Leaving"]
graydon has left #ocaml []
graydon has joined #ocaml
<timmy> when i have a try..with.. inside of a match abc with ..., how do i get the next x -> y in the match set to not be confused with the previous x -> y where the y has a try..with.. at the end of it
thelema|away is now known as thelema
thelema is now known as thelema|away
IRCMonkey has joined #ocaml
IRCMonkey has quit ["ChatZilla 0.8.7 [Mozilla rv:1.0rc3/20020523]"]
IRCMonkey has joined #ocaml
gl has quit [Read error: 113 (No route to host)]
* Yurik is back (gone 19:14:30)
<Yurik> guys, is there any reflection facilities in OCaml? I mean, for example, can I inspect module sig or module struct within ocaml program?
malc has joined #ocaml
skylan has quit [Read error: 104 (Connection reset by peer)]
skylan has joined #ocaml
stefcon has joined #ocaml
<Yurik> is there anobody using tuareg w/ symlock?
stefcon is now known as Owll
<Owll> hia
<Yurik> Owll: hi
<Owll> nope :)
timmy has quit [Read error: 110 (Connection timed out)]
Owll has quit [Read error: 113 (No route to host)]
timmy has joined #ocaml
malc has quit ["no reason"]
TimFreeman has joined #ocaml
TimFreeman has left #ocaml []
__DL__ has joined #ocaml
__DL__ has quit ["Client Exiting"]
thelema|away is now known as thelema
gluck has quit ["leaving"]
exa has joined #ocaml
<exa> hi!
<exa> can I read floats, doubles or 64 bit integers from a binary file in ocaml?
<exa> i looked at the docs but couldn't figure it out
<thelema> exa: what docs did you look at?
<exa> library docs
<exa> i might be too much of a newbie to have skipped it
<exa> i mean i did see functions
<thelema> which library?
<exa> that let you read integers from a binary file, right?
<exa> but for other 'integral' types?
<thelema> what you'll probably need to do is read bits in from a file into an integer and then convert that bit-integer into the appropriate type.
<thelema> val float_of_bits : int64 -> float
<exa> hah thanks
<exa> so i read 8 bytes, and use that function then
<exa> cool
<exa> hmm i want to write a machine learning library in ocaml
<exa> maybe it lets me try things faster ;)
<exa> binary files are cruicial. lots of arcane file formats.
<thelema> exa: what paradigms? neural nets and decision trees come to mind, but what else?
<exa> my expertise is frequency mining
<exa> i basically want to specify the interface and leave the implementation to people
<exa> and implement a few of my own
<exa> but yes, decision trees would be good as well as NNs
<exa> like C4.5
<exa> and multi-layer FF nets with BP
<exa> would be great
<thelema> frequency mining... I can try to guess what that's about...
<exa> first phase of association rule mining
<exa> where you find sets of items that occur frequently in a transaction phase
<exa> hm i can use that phrase in the abstract heh
<exa> thelema: are you interested in machine learning?
<exa> transaction phase -> transaction database
<exa> or transaction set
<exa> sorry
<thelema> I've done my fair share of work in that area, but have had to leave it aside for about a year while I do other things.
<exa> hm great
<exa> i'm trying to predict the stock market now
<exa> it turns out to be quite predictable
<exa> :)
<exa> maybe we can get rich, who knows? :)
<thelema> good luck with the stock market thing.
<exa> heh
<exa> thank you
<exa> it's fun
<exa> i just finished a new estimator in haskell
<exa> but I think I might have to drop it in favor of ocaml one day
<exa> haskell isn't good for efficiency
<exa> i mean when it comes to doing things iteratively, or processing large amounts of data
<exa> I didn't find it good enough
<thelema> ocaml is a nice language. it seems to get the best of a lot of worlds.
<exa> great type system, and still efficient
<exa> i like the module system, too
<exa> the objective features look nice, too
<exa> I wanna write a binding for Qt library one day
<exa> And KDE libraries
<thelema> I really like the type system... it's just got the right types.
<exa> yea, i'm a newbie now but it's easy after haskell :)
<thelema> And exceptions that carry data back with them are just the best.
<exa> yea like C++
* thelema doesn't program in C++
<thelema> juct C
<thelema> *just
IRCMonkey has quit [Read error: 110 (Connection timed out)]
yond has joined #ocaml
kjs3 has joined #ocaml
thelema is now known as thelema|away
exa has quit ["KVIrc 2.1.2-pre2 'Monolith'"]
gl has joined #ocaml
yond has quit ["ChatZilla 0.8.7 [Mozilla rv:1.0rc3/20020523]"]
<timmy> i'm having a problem with typing in ocaml
<timmy> method print_blists stream id =
<timmy> self#print_items stream id [1] (fun x y -> true)
<timmy> (Array.to_list blists) (fun x -> string_of_int x.blist_id)
<timmy> method print_ilists stream id =
<timmy> self#print_items stream id [1] (fun x y -> true)
<timmy> (Array.to_list ilists) (fun x -> string_of_int x.ilist_id)
<timmy> This expression has type ilist list but is here used with type blist list
<timmy> i'm trying to get the function to work on all types of lists
thelema|away is now known as thelema
kjs3-- has joined #ocaml
kjs3 has quit [Read error: 110 (Connection timed out)]
kjs3-- is now known as kjs3
graydon has quit [Read error: 113 (No route to host)]
gl has quit [Read error: 104 (Connection reset by peer)]
gl has joined #ocaml
kjs3 has quit [Read error: 110 (Connection timed out)]
<smkl> timmy: in current version of ocaml, methods cannot be polymorhic
gl has quit [Read error: 113 (No route to host)]
gl has joined #ocaml
* gl is away: code
<timmy> smkl: so regular functions can be?
<thelema> I don't think so.
<timmy> hmm
<smkl> regular functions can be polymorphic. but perhaps you can design your methods differently
<thelema> smkl: not very polymorphic
<timmy> i think i got it to compile at least
<smkl> regular functions are HM polymorphic, that's almost always enough
<thelema> HM?
<smkl> hindley-milner
<thelema> I'll have to look that up sometime.
<smkl> that just means that all type variables are bound at the highest level ... for example ('a -> 'b) -> 'a -> 'b means "all 'a. all 'b. ...". in system F they could be bound in other ways
<thelema> fair enough. hindley-milner polymorphism seems like the most common form.
<thelema> it basically means you can have functions with unbound types. that work on any type you plug in.
bobacus has joined #ocaml