MS-Windows compiler review and Ada 'myopia'

[emery@mitre-bedford.arpa (David Emery)]

The latest issue of IEEE Computer (June 93, pp 108-111) has a review
of two Ada compilers (Meridian and Janus).  The author, Frank Pappas,
complained about the Windows API thin bindings included with the two
compilers.  He said, "...I think it was a mistake for [the compiler
vendors] to hav deefined their own interfaces.  What's needed is a
common, high-level Ada binding for Windows, which will probably come
about only if someone at the Department of Defense is willing to pay
for it - something that should be done soon!"

Well, I have several problems with this.  As Mike Feldman and Greg
Aharonian, among others, have pointed out, a problem with the Ada
business is that they seem to wait for the DoD to pay for everything.
I agree that a 'common' MS-Windows API binding would be a very good
thing, but I don't think that we should sit back and wait until "Uncle
Sugar" pays for it.  If there's sufficient interest, then certainly
the vendors and interested users can get together and develop a
'standard' Windows Ada binding.  I'm sure that ACM SIGAda would be
willing to sponsor a working group to facilitate this discussion
(although ACM does not develop standards, so the 'product' of such a
group would need some other mechanism for formal standardization.) 

Besides, I don't agree that it is in the DoD's interests to fund the
Ada binding to MS-Windows.  Windows is a very proprietary API, owned
by MicroSoft.  The current DoD trend is away from such proprietary
interfaces, and towards open, standard interfaces.  I think it is much
more appropriate (and important) for DoD to sponsor X Windows, etc.
bindings, and in particular a binding to IEEE draft 1295.1 (MOTIF).
With respect to MS-Windows, it'll be interesting to watch the Sun
effort to develop a 'standard' Windows API that is not proprietary to
MicroSoft. 

				dave