Hello all, I opened a thread some time ago where I asked whether it's enough to have only immediate switching. But it dawned on me that Reactive switches are actually all delayed, since even though the delay is infinitesimal in principle, we still have to wait for the next point of observation until we can see its effect. This is a huge problem in practice, because this way every stateful signal imposes a delay equal to the length of the period between observations. Here's a simple test case for illustration:
import Control.Applicative import FRP.Reactive
tick :: Event () tick = atTimes [0..]
behs :: [Behavior Double] behs = pure 1 : map (integral tick) behs
eventList :: Event a -> [a] eventList e = x : eventList e' where (x,e') = firstRestE e
testBeh :: Behavior a -> [a] testBeh b = take 10 $ eventList (snapshot_ b tick)
test :: IO () test = mapM_ (print . testBeh) (take 5 behs)
Running test prints the following: [1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0] [0.0,0.0,1.0,2.0,3.0,4.0,5.0,6.0,7.0,8.0] [0.0,0.0,0.0,1.0,3.0,6.0,10.0,15.0,21.0,28.0] [0.0,0.0,0.0,0.0,1.0,4.0,10.0,20.0,35.0,56.0] [0.0,0.0,0.0,0.0,0.0,1.0,5.0,15.0,35.0,70.0] I ran into the very same problem when working on the first version of Elerea. Transfer functions in version 0.1.0 behave exactly the same way, and I switched to immediate transfer functions from 0.2.0. The difference is obvious if you look at the breakout example in action: in 0.1.0, the ball takes long to react to collisions, since the dependency cycle consists of three items (bricks, velocity, position), so any change takes just as many frames to propagate. From 0.2.0 onwards, collisions are much more accurate without any change in the example code. Sure, it is possible to work around the delay by combining the switching value with the one that caused its switch, but this adds a lot of complexity to the code on the user end. You can make life easier by introducing immediate versions as helper functions, but then you just reintroduced this distinction. Of course this is not surprising, since neither delayed nor immediate switching is sufficient on its own. But this means that Reactive should support both out of the box and not encourage ad hoc workarounds that are likely to break in unexpected ways. Also, immediate switching seems the more sensible default, because it's more straightforward to derive the delayed version from it than the other way around, and it is what we need most of the time. Are there any plans to address this problem? Or a completely new angle to look at it from? Gergely -- http://www.fastmail.fm - Faster than the air-speed velocity of an unladen european swallow