diff --git a/Numeric/FloatMinMax.hs b/Numeric/FloatMinMax.hs
new file mode 100644
--- /dev/null
+++ b/Numeric/FloatMinMax.hs
@@ -0,0 +1,38 @@
+-- | The minimum and maximum positive, finite floats.
+
+module Numeric.FloatMinMax
+( floatMin
+, floatMaxDenorm
+, floatMinNorm
+, floatMax
+)
+where
+
+import Control.Applicative
+import Data.Function
+
+-- | The minimum positive, denormalized float.
+floatMin :: RealFloat a => a
+floatMin = fix $ do d     <- floatDigits
+                    (e,_) <- floatRange
+                    pure $ encodeFloat 1 (e-d)
+{-# INLINE floatMin #-}
+
+-- | The maximum denormalized float.
+floatMaxDenorm :: RealFloat a => a
+floatMaxDenorm = floatMinNorm - floatMin
+{-# INLINE floatMaxDenorm #-}
+
+-- | The minimum positive, normalized float.
+floatMinNorm :: RealFloat a => a
+floatMinNorm = fix $ do (e,_) <- floatRange
+                        pure $ encodeFloat 1 (e-1)
+{-# INLINE floatMinNorm #-}
+
+-- | The maximum finite float.
+floatMax :: RealFloat a => a
+floatMax = fix $ do r     <- floatRadix
+                    d     <- floatDigits
+                    (_,e) <- floatRange
+                    pure $ encodeFloat (r^d-1) (e-d)
+{-# INLINE floatMax #-}
diff --git a/zoom-cache.cabal b/zoom-cache.cabal
--- a/zoom-cache.cabal
+++ b/zoom-cache.cabal
@@ -7,7 +7,7 @@
 -- The package version. See the Haskell package versioning policy
 -- (http://www.haskell.org/haskellwiki/Package_versioning_policy) for
 -- standards guiding when and how versions should be incremented.
-Version:             0.1.0.0
+Version:             0.1.0.1
 
 -- A short (one-line) description of the package.
 Synopsis:            A streamable, seekable, zoomable cache file format
@@ -56,6 +56,7 @@
                        Data.ZoomCache.Summary
                        Data.ZoomCache.Write
                        Data.Iteratee.ZoomCache
+                       Numeric.FloatMinMax
   
   -- Packages needed in order to build this package.
   -- Build-depends: