Keean, As far as I can tell, both your solutions to the "one-shot" problem require that: (a) the expression to be one-shotted is in the IO monad. That seems reasonable, since why else does one care (semantically speaking)? (b) they depend on the host operating system platform (semaphores, process id, environment variables) rather than pure Haskell language features. Wouldn't it be easier to simply define "once" as a common Haskell library function? #g -- At 23:36 10/11/04 +0000, Keean Schupke wrote:
Hi, Here's another completely safe (and simpler way) to limit a computation to only happen once:
once' :: IO () -> IO () once' f = do k <- getProcessID a <- getEnv (showString "MyApp.Main" $ show k) case a of Just _ -> return () _ -> do f setEnv (showString "MyApp.Main" $ show k) "" False
Actually both this and the semaphore example show that there is probably an alternative to allowing top-level '<-' type definitions - and that would be to have named-MVars in Haskell. It would be quite easy to code these as a small C library - and then FFI import the bindings to Haskell. I don't know whether from a 'purists' point of view whether this represents anything better than module-initialisations - but it does remove the diamond-inheritance style problem you get (if B imports A and C imports A and Main imports B and C, does A's init get run twice? are there two copies of the variables, each initialised once, or one copy that gets initialised twice?). But either way the idea could be implemented without requiring changes to the language spec or the compilers, and would just be a library which you could use.
Keean.
_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
------------ Graham Klyne For email: http://www.ninebynine.org/#Contact