I was under the impression that STM code needed to be in its own monad. I was looking at Control.Concurrent.STM.TChan, for example, where signatures like this exist: newTChan :: STM (TChan a) readTChan :: TChan a -> STM a writeTChan :: TChan a -> a -> STM () and then newTChan :: STM (TChan a) readTChan :: TChan a -> STM a writeTChan :: TChan a -> a -> STM () I guess I should give this another look, re-read the STM paper and check out your patch. Regardless, simple is elegant and your Maybe solution is simple. Thanks, Joel On Nov 22, 2005, at 7:09 AM, Tomasz Zielonka wrote:
I am talking about Software Transactional Memory, which is in Control.Concurrent.STM. I think you confused it with State Transformer Monad.
In your case STM would allow you to wait simultaneously on (T)MVar and (T)Chan. It would look like this:
logger :: TMVar () -> IO () logger die = join $ atomically $ (do x <- readTChan parent return $ do putStrLn x logger die) `orElse` (do takeTMVar die return (return ()))