Re: Seeking for papers about tagged types vs access to subprograms

[Robert A Duff <bobduff@shell01.TheWorld.com>]

"Randy Brukardt" <randy@rrsoftware.com> writes:

>...Our experience has been that we introduce about as 
> many serious bugs to the standard as we make changes to it, simply because 
> we can't easily tell when we break things.

I don't think it's QUITE that bad.  I think there were around 1000
AI's for Ada 83, 400 for Ada 95, and 300 for Ada 2005.  Not all of
those were approved.  I suspect the vast majority of Ada 83 AI's
were bug reports, whereas for Ada 95 and 2005, many were enhancements.
So the number of bug reports has been going down over the years.

But your point is correct:  We can't do automated regression testing
on the RM, so it's easy to insert bugs.

I don't think it's possible to do a major overhaul of the Ada type
system while remaining compatible.  But then, in 1990 or so, I
didn't think it was possible to add object-oriented support to
Ada 83.  Tucker proved me wrong.  ;-)

- Bob