Hello
Here is the latest Caml Weekly News, for the week of February 13 to 20, 2007.
> let me point out that amb is supposed to work as an _angelic_
> nondeterministic choice operator. This means it must choose a value
> that, if at all possible, leads to successful completion of the
> computation.... The scheme implementation involves callcc magic.
> If anyone knows a reasonable implementation of amb, I'd be
> interested to know.
Here it is. Ocaml has something better than call/cc: delimited
continuations. So, amb is trivially implementable, in two lines of
code. We also need a `toplevel function', to tell us if the overall
computation succeeded. One may think of it as St.Peter at the
gate. For now, we take a computation that raises no exception as
successful. In general, even non-termination within a branch can be
dealt with intelligently (cf. `cooperative' threading which must yield
from time to time).
Regarding effects incurred while evaluating branches: one deal with
them as one deal with effects when implementing any transactional
mechanism: you prohibit them, you log the updates, you log the state
before the beginning so to undo, you use zipper for functional
`mutations' (which can be done with delimited continuations, too), you
operate in a sandbox, etc. The ZFS talk gives an example:
http://pobox.com/~oleg/ftp/Computation/Continuations.html#zipper-fs
Here are the tests. The second one requires a three-step clairvoyance:
let test1 () =
let v =
if (amb [(fun _ -> false); (fun _ -> true)]) then
7
else failwith "Sinner!"
in Printf.printf "got the result %d\n" v;;
let test1r = toplevel test1;;
(* got the result 7 *)
(* Test that invokes the Pythagorean spirit *)
let numbers = List.map (fun n -> (fun () -> n)) [1;2;3;4;5];;
let pyth () =
let (v1,v2,v3) =
let i = amb numbers in
let j = amb numbers in
let k = amb numbers in
if i*i + j*j = k*k then (i,j,k) else failwith "too bad"
in Printf.printf "got the result (%d,%d,%d)\n" v1 v2 v3;;
let pythr = toplevel pyth;;
(* got the result (3,4,5) *)
(* Open the DelimCC library
http://pobox.com/~oleg/ftp/Computation/Continuations.html#caml-shift
*)
open Delimcc;;
let shift p f = take_subcont p (fun sk () ->
push_prompt p (fun () -> (f (fun c ->
push_prompt p (fun () -> push_subcont sk c)))))
;;
(* How evaluation has finished *)
type res =
Done (* Finished with the result *)
| Exc of exn (* Got an exception -- no good *)
| Choices of (unit -> res) list (* Alternative universes *)
exception Amb (* Raise when all choices are bad *)
;;
let topprompt = new_prompt ();;
(* If it looks like an OS scheduler, it's because it is *)
let toplevel thunk =
let rec loop queue = function
| Done (* evaluation of a branch finished successfully *)
-> ()
| Exc _ -> try_another queue
| Choices more -> (* OK, add them to the end: breadth-first *)
try_another (queue @ more)
and try_another = function
| [] -> raise Amb (* No more choices *)
| h::t -> loop t (try h () with e -> Exc e)
in
loop [] (push_prompt topprompt
(fun () -> try let _ = thunk () in Done with e -> Exc e))
;;
(* Split the universe. Something like `fork (2)' *)
let amb choices = shift topprompt (fun sk ->
Choices (List.map (fun choice -> (fun () -> sk choice)) choices));;
I've always thought it would be nice if it were possible to use OpenGL calls in a window created with the Ocaml Graphics module. This would prevent the need to require external toolkits such as SDL or GLUT for some applications. The Ocaml Graphics module is quite limited compared to SDL or GLUT, and also fills a slightly different niche, but it does come with a number of useful functions for drawing text, various graphics primitives, and for polling mouse and keyboard events. Simple standalone programs or demos can be created and distributed without requiring that the SDL or GLUT shared library be present on those systems. To get an Ocaml window working with OpenGL I wrote a small stub file in C and a few lines of ML code. A gzipped tar file containing all necessary code to get OpenGL working in a native Ocaml graphic window, along with an example for windows and linux, can be obtained from http://www.elliottoti.com/code/glgraphics/glgraphics.tar.gz A screenshot and a little more documentation is available at http://www.elliottoti.com/index.php?p=24
Here is a quick trick to help you read this CWN if you are viewing it using vim (version 6 or greater).
:set foldmethod=expr
:set foldexpr=getline(v:lnum)=~'^=\\{78}$'?'<1':1
zM
If you know of a better way, please let me know.
If you happen to miss a CWN, you can send me a message and I'll mail it to you, or go take a look at the archive or the RSS feed of the archives.
If you also wish to receive it every week by mail, you may subscribe online.