On Thu, 5 Jun 1997, Giles Lean wrote:

> 
> > extensions), but has been tweaked so that errors which would have
> > occurred at compile time under K&R C will now occur at run time.
> 
> The concerns here are valid, but practically theoretical.  I've never
> seen a type promotion bug actually cause damage.  Do people actually
> see such problems "in the field"?

Well, several examples have been pointed out of functions which take
char types as arguments, a situation which promotes differently in ANSI
and K&R C.  We have been saved a lot of the hassle of this because of
the fact that for all the work we put into supporting K&R C, we don't
actually use it, and because of the gcc extensions we use, _couldn't_
use it on any compiler save gcc, which uses ANSI promotion rules even in
K&R C...  (mostly)

I'm much more concerned with the unreliability introduced (and the
time wasted) in the process of backporting code to K&R C, a process
which seems to have no actual utility...

--
				Jim Wise
				jim.wise@turner.com