For what it's worth, let me throw a few comments into this mix.
Allchin's argument is, essentially, that the government should stop
funding any work that would be released under GPL-type restrictions.
Effectively, if it were to do so, the government would be preferring
the proprietary model to the GPL-type open-source one. I'm not going
to add to the heated debate of the merits of one model over the other.
I think most of us would agree that _both_ models have their strengths
and their weaknesses. This debate has been useful in clarifying that
issue. I hope you would agree that it would be foolish for Uncle Sam
to choose a winner between the two.
Specifically on Ada 95 (after all, this _is_ Team-Ada!), Uncle Sam
has invested roughly equally in both models. Thanks to AJPO, we
have both GNAT and ObjectAda. As we know, GNAT is GPL-ed and
OA is proprietary.
Undoubtedly you all know that Uncle Sam funded the original NYU-based
work on GNAT.
Perhaps you're not all aware that the original ObjectAda front end was
developed by Intermetrics in the guise of the AJPO-sponsored "Academic
Ada" project; Thomson (later Aonix) became the commercializer thereof.
If memory serves, the U.S. taxpayers' investment was in the
neighborhood of a few million dollars in each system; this teeny bit
of seed money has given rise to successful projects under both models.
(Well, successful as Ada goes, anyway. :-))
In the interest of full disclosure: one of the competitors for the
Academic Ada contract was a team that I was putting together, with
work at GW and participation from the then-new ACT and a couple of
other small companies. GNAT would have been the compiler in the
middle of an integrated development environment. The whole would
have been GPL-ed, of course.
As angry as I was at losing the award, with benefit of hindsight,
I can see that AJPO may have done the right thing by balancing
its investment between the two models; their doing so gave Ada 95 a
kick start that neither model, by itself, could have done.
Bottom line: Uncle Sam shouldn't pick winners and losers. Let them
fund GPL-ish software, public-domain software, and maybe even
proprietary software with a reasonably evenhanded balance, on
the merits of each proposal they get.
BTW - I would support a similarly evenhanded balance in
funding R&D in other industries too, especially in pharmaceuticals.
In the drug industry, I opine that the scales are tipped too heavily
toward the proprietary model - the taxpayers fund a lot of the R&D,
the drug industry patents and commercializes the result and then claims
their outrageous U.S. prices are necessary to fund their own R&D.
Actually, I think most of that outrageous price is paying for
the endless TV ads we are seeing for expensive prescription drugs.