Re: What is differance between thick and thin binding?

[tmoran@bix.com]

  There was a thread on this some months ago (check dejavu) where
somebody gave a nice, succinct, explanation.  There's a much more
elaborate discussion in our TriAda paper on Claw, available
at www.rrsoftware.com
  A thin binding retains the style of the original.  It has just
enough changes to make it legal Ada, instead of C (or whatever).
A very thin binding can largely be generated by a program.
  A thick binding has an Ada style - it uses overloading,
exceptions instead of function return values, Ada-ish names, type
checking, Controlled records, etc.  and generally looks like
something an Ada programmer would have designed.  The
abstraction level is normally much higher than the thin
binding and a lot of the details have already been handled.
  There are of course different degrees of thickness.  For
instance, Claw contains the Windows message loop internally,
hiding all sorts of messy details, but the procedure
Claw.Edit.Empty_Undo_Buffer merely encapsulates sending an
EM_EMPTYUNDOBUFFER message to the window.  A really thick
binding, such as Java or Tcl/Tk, can present the same abstraction
across different platforms, but at the cost of making it
difficult to get a "Windows look and feel".  You pays your
money and you takes your choice.