Re: Meridian Ada for MS Windows

Re: Meridian Ada for MS Windows

[David Emery]

This is an instance of the larger debate on "thin vs thick" language
bindings.  

The arguments for thin bindings are:

	1.  efficiency (no extra Ada layering)
	2.  reuse of existing documentation
	3.  ease of development

The arguments for thick bindings are:

	1.  appropriate Ada coding style/abstraction
	2.  support for Ada features (e.g. exceptions, tasking)

The POSIX/Ada binding (IEEE Standard P1003.5) is an example of a thick
binding.  This document has a lot of (very good) rationale on the
topic.

Personally, I'm very much in favor of thick bindings as products.  If
all I wanted was a thin binding, I'd either program in C, or do it myself.

				dave

Re: Meridian Ada for MS Windows

[Bob Kitzberger]

emery@Dr_No.mitre.org (David Emery) writes:

>The arguments for thin bindings are:
>
>	3.  ease of development

This strikes me as an argument for _thick_ bindings, not thin.  Why
would thin bindings ease development over thick bindings?

	.Bob.
----------------
Bob Kitzberger          VisiCom Laboratories, Inc.
rlk@visicom.com         10052 Mesa Ridge Court, San Diego CA 92121 USA
                        +1 619 457 2111    FAX +1 619 457 0888

Re: Meridian Ada for MS Windows

[David Emery]

Thin bindings are easier for the binding developer to produce.  
(IMHO, they're not easier for the applications developer to use.
 They're easier for the binding developer becaue they remove his
 requirement to think about what he's doing...)

				dave