On Fri, Dec 30, 2016 at 05:16:21PM +0100, Ertugrul Söylemez wrote:
could you show us the source code of your event wires? I suspect that this is not a Netwire issue at all, but an issue of one of the event wires draining the MVar, so the other one never actually sees a full MVar.
Hello Ertes, the source code is attached to the original message [1]; I'll paste here the relevant bits. This is the wire where the events are generated:
keyEvent :: Wire s () IO (MVar Char) (Event Char) keyEvent = mkGen_ f where f :: MVar Char -> IO (Either () (Event Char)) f r = tryTakeMVar r >>= \mmv -> case mmv of Nothing -> return $ Right NoEvent Just c -> putStrLn ("pop: " ++ [c]) >> return (Right (Event c))
And these are the three wires which filter/merge `keyEvent`(s):
keyA, keyQ :: Wire s () IO (MVar Char) (Event Char) keyQ = filterE (== 'q') . keyEvent keyA = filterE (== 'a') . keyEvent
testWire2 :: Wire s () IO (MVar Char) Char testWire2 = hold . (keyQ &> keyA)
`testWire2` is the function that is giving me headaches (works for Q keypresses but not A ones). [1] https://mail.haskell.org/pipermail/beginners/attachments/20161227/c0585a78/a...