Re: OO and "mainframes"

James Gavan wrote:

> Bit too late for DOD to get involved again - but IF they were still
> involved, and accepting ISO(WG4) 'calls the shots' - DOD involvement
> logically shouldn't put anybody's nose out of joint. ISO gives the
> imprimatur, but anybody wishing to sell to DOD has to fully comply with
> the ISO Standard. Implication of that - we would probably see an
> increase in membership of J4 - so that current vendors not on the list
> have a chance to 'get their oar in'.

Actually, I wouldn't care whether it's DOD, GSA, DIN or ITSCJ that  made
such a requirement -- the point right now is that there's no requirement.

There's a chicken-and-egg problem here, too, and that has to do with the
state of the validation suite:  I know for certain that no full 2002
validation suite has been developed, and so far as I know the validation
suite that *does* exist for the '85 standard doesn't even include provisions
for the Intrinsic Function amendment of 1989, much less the Corrections
Amendment of 1994.

The imposition of a requirement that vendors validate their compilers
against a suite of programs has meaning only if such a suite of programs
exists before the requirement is in force, and that suite does not currently
exist (at least for the new features in 2002 COBOL).  Barring a volunteer
effort on the part of the COBOL community to provide the revisions to the
current '85 validation suite, it pretty much falls to any agency demanding
compliance to undergo the expense of ensuring that a means of validation was
available, and that obligation forms a significant disincentive to making
the requirement in the first place.

-Chuck Stevens