Re: GCC 3.1 released

["Jeffrey Creem" <jeff@thecreems.com>]

"Robert Dewar" <dewar@gnat.com> wrote in message
news:5ee5b646.0205181220.66cba6c2@posting.google.com...
> "Jeffrey Creem" <jeff@thecreems.com> wrote in message
news:<PmtF8.29827$sg2.7831839@typhoon.ne.ipsvc.net>...
>
> > Yup..There are problems.. There are also problems in GNAT
> > 3.14, 3.13, 3.12, 3.15 ... etc.
>
> But none of these releases failed ACATS tests and tests in
> our test suite. GNAT 5 has significant numbers of failures
> in all categories on all targets. Several of these are
> of the form of incorrect code being generated silently
> and resulting in wrong results. I am not saying that the


And of course I am not going to argue with Robert about
the quality of any particular GNAT release since that would be somewhat
silly.

But... The fact that past public releases pass all ACATS tests and
3.1 fails some and can produce incorrect code is somewhat bad
however I strongly suspect that at some point in their life
cycles each of those public releases produced incorrect
code on some customers code somewhere.

Passing ACATS does give me a warm fuzzy but regression
testing any new compiler release on my own code base is the
only thing that really convinces me of the quality of any given
compiler for my own use.

Still, I would not suggest that people use 3.1 for some important
project with a real customer but then again I don't think it is really
a great idea to use any of the other versions (of any complicated
product) for that purpose without support either.

Disclaimer : Of course I have never had the opportunity of actually using
an ACT released compiler on any project....  I have however used
plenty of support compilers that have passed various conformance suites and
still been "broken".

P.S. I am sure I said something above that conflicts with my first statement
about not arguing with Robert about GNAT... I am glad I just said it
was a silly idea and not a stupid one :)