o 0.19
   o External content qualifiers are cleared now if main qualifiers
     go 'up' now. Maybe this fixes contexts created by lone events
     like MCP does (it sucks!).
   (Chris)

 o 0.18
   o External mouse events did not carry the qualifiers of the main
     input devices (i.e. shift-clicking with the usb mouse was not
     possible). Fixed.
   (Chris)

 o 0.17
   o Now only a few events actually create qualifier context. This
     should fix the famous shift-bug or the console if that was
     the problem. Otherwise I dunno.
   (Chris)

 o 0.16
   o fixed ConCat loop bug: ie_Code was not checked to be the same,
     therefore mouse up events could have been eaten up.
   o PeekQualifier() does not return all context qualifiers anymore,
     but leaves out the context qualifiers of the calling task (if
     any). This should fix some problems which merge the
     PeekQualifier() value to the events written back to the device.
   (Chris)

 o 0.15
   o lots of changes to allow 68k compiles.
   o added context stuff for WriteEvent. Tasks that use WriteEvent
     get a context that keeps track of the qualifiers which will
     be added globally (and to PeekQualifier). RawKey events are
     even passed through the internal key management routine, so
     that key repeat is supported correctly. So in theory, I could
     deactivate all patching in FormAldiHyd now. USB mice and
     keyboards work like a charm.
   (Chris)

 o 0.14
   o improved repeat mode where it didn`t cancel a repeat
     of one key if a different key was released
     (Ralph)

 o 0.13
   added retargable input device support
   o IO_GetKeyBoard
   o IO_SetKeyBoard
   o IO_GetMouse
   o IO_SetMouse
   (Ralph)

 o 0.12
   fixed abortio problem
   (Ralph)

 o 0.11
   Fixed qualifier repeat problem..now for keypads too:-)
   (Ralph)

 o 0.10
   Fixed qualifier repeat problem
   (Ralph)

 o 0.9
   Increased stacksize
   (Ralph)

 o 0.8
   Added new GET#? IO commands
   (Ralph)

 o 0.7
   fixed intuiticks problem
   added IND_SETINTUITICKS cmd...because the problem`s effect
   on the UI was liked by some people:-)
   (Ralph)

 o 0.6
   Worked around Vinced bugs.
   1. OpenDevice("input.device" with a Unit 1 alias UNIT_VBLANK
   2. A failure here results into a bogus FreeVecPooled which
      then results into a custom alert.
   3. Use of semaphores in inputhandlers..bad code.
   (Ralph)

 o 0.5
   Removed strict-align option...crap code
   (Ralph)

 o 0.4
   fixed single key qualifier events
   (Ralph)

 o 0.3
   fixed the key qualifier problem during mouseevents
   (Ralph)

 o 0.2
   experimental changes
   (Ralph)

 o 0.1
   First version
   (Ralph)