diff --git a/CHANGELOG b/CHANGELOG
--- a/CHANGELOG
+++ b/CHANGELOG
@@ -1,3 +1,9 @@
+2018-08-11 Ivan Perez <ivan.perez@keera.co.uk>
+        * Yampa.cabal: Version bump (0.11.1).
+        * README.md: Documents papers.
+        * src/: Fixes leak.
+        * Thanks to @tresormuta, @chriz-keera.
+
 2018-04-05 Ivan Perez <ivan.perez@keera.co.uk>
         * Yampa.cabal: Version bump (0.11).
         * src/: Adds documentation; makes type synonym a newtype.
diff --git a/Yampa.cabal b/Yampa.cabal
--- a/Yampa.cabal
+++ b/Yampa.cabal
@@ -1,5 +1,5 @@
 name: Yampa
-version: 0.11
+version: 0.11.1
 cabal-version: >= 1.8
 license: BSD3
 license-file: LICENSE
diff --git a/src/FRP/Yampa/InternalCore.hs b/src/FRP/Yampa/InternalCore.hs
--- a/src/FRP/Yampa/InternalCore.hs
+++ b/src/FRP/Yampa/InternalCore.hs
@@ -534,22 +534,46 @@
 
 -- | Choice of which SF to run based on the value of a signal.
 instance ArrowChoice SF where
-    left sf = SF $ \a ->
-                     -- NOTE: there might be a problem with choice here.
-                     -- Do the delta times accumulate for the unused branch?
-                     -- Recommendation by Olivier Charles: take a look
-                     -- at Settable Signals paper, it discusses which
-                     -- option would be best.
-                     case a of
-                       Left x  -> let (sf', b') = sfTF sf x
-                                  in (futureArrowLeft sf', Left b')
-                       Right x -> let sf' = SF' $ \_ -> sfTF sf
-                                  in (futureArrowLeft sf', Right x)
-       where futureArrowLeft fSF = SF' $ \dt a ->
-                case a of
-                  Left x  -> let (sf', b') = sfTF' fSF dt x
-                             in (futureArrowLeft sf', Left b')
-                  Right x -> (futureArrowLeft fSF, Right x)
+  -- (+++) :: forall b c b' c' . SF b c -> SF d e -> SF (Either b d) (Either c e)
+  sfL +++ sfR = SF $ \a ->
+    case a of
+      Left b  -> let (sf', c) = sfTF sfL b
+                 in (chooseL sf' sfR, Left c)
+      Right d -> let (sf', e) = sfTF sfR d
+                 in (chooseR sfL sf', Right e)
+
+    where
+
+      -- (+++) for an initialized SF and an SF
+      --
+      -- chooseL :: SF' b c -> SF d e -> SF' (Either b d) (Either c e)
+      chooseL sfCL sfR = SF' $ \dt a ->
+        case a of
+          Left b  -> let (sf', c) = sfTF' sfCL dt b
+                     in (chooseL sf' sfR, Left c)
+          Right d -> let (sf', e) = sfTF sfR d
+                     in (choose sfCL sf', Right e)
+
+      -- (+++) for an SF and an initialized SF
+      --
+      -- chooseR :: SF b c -> SF' d e -> SF' (Either b d) (Either c e)
+      chooseR sfL sfCR = SF' $ \dt a ->
+        case a of
+          Left b  -> let (sf', c) = sfTF sfL b
+                     in (choose sf' sfCR, Left c)
+          Right d -> let (sf', e) = sfTF' sfCR dt d
+                     in (chooseR sfL sf', Right e)
+
+      -- (+++) for initialized SFs
+      --
+      -- choose :: SF' b c -> SF' d e -> SF' (Either b d) (Either c e)
+      choose sfCL sfCR = SF' $ \dt a ->
+        case a of
+          Left b  -> let (sf', c) = sfTF' sfCL dt b
+                     in (choose sf' sfCR, Left c)
+          Right d -> let (sf', e) = sfTF' sfCR dt d
+                     in (choose sfCL sf', Right e)
+
 
 
 -- | Signal Functions as Arrows. See "The Yampa Arcade", by Courtney, Nilsson
diff --git a/src/FRP/Yampa/Simulation.hs b/src/FRP/Yampa/Simulation.hs
--- a/src/FRP/Yampa/Simulation.hs
+++ b/src/FRP/Yampa/Simulation.hs
@@ -17,7 +17,7 @@
 -- called /reactimation/, and the latter is called /embedding/.
 --
 -- Normally, to run an SF, you would use 'reactimate', providing input samples,
--- and consuming the putput samples in the 'IO' monad. This function takes over
+-- and consuming the output samples in the 'IO' monad. This function takes over
 -- the program, implementing a "main loop". If you want more control over the
 -- evaluation loop (for instance, if you are using Yampa in combination with a
 -- backend that also implements some main loop), you may want to use the
@@ -125,7 +125,7 @@
 reactimate :: Monad m
            => m a                             -- ^ Initialization action
            -> (Bool -> m (DTime, Maybe a))    -- ^ Input sensing action
-           -> (Bool -> b -> m Bool)           -- ^ Actuaction (output processing) action
+           -> (Bool -> b -> m Bool)           -- ^ Actuation (output processing) action
            -> SF a b                          -- ^ Signal function
            -> m ()
 reactimate init sense actuate (SF {sfTF = tf0}) =
diff --git a/src/FRP/Yampa/Switches.hs b/src/FRP/Yampa/Switches.hs
--- a/src/FRP/Yampa/Switches.hs
+++ b/src/FRP/Yampa/Switches.hs
@@ -51,7 +51,7 @@
 --'Functor' or a list ('[]').
 --
 -- - How the input is router /(B\/Z\/ )/: when multiple SFs are being combined,
--- a decision needs to be made about how the input is passed ot the internal
+-- a decision needs to be made about how the input is passed to the internal
 -- SFs.  In some cases, broadcasting is used to pass the same input to all
 -- internal SFs. In others, the input is itself a collection, and each element
 -- is passed to one internal SF (i.e., /zipping/). In others, an auxiliary