> I wouldn't object to a proposal under which an old-style definition
> behaved like a prototype, but many of the members of the committee
> would, so we probably won't get one.  :(

If this:

#ifdef __STDC__
#define _P(x) x
#else
#define _P(x)
#endif

int main _P ((int, char **, char **));

...

int main (argc, argv, envp)
	int argc;
	char **argv;
	char **envp;
{
...
}

...is no longer legal, I will have a serious bone to pick.   As
maintainer of the Internet Software Consortium DHCP server, I provide
support for a number of legacy systems that _do_not_ have ANSI C
compilers and never will, because they are not being updated.   SunOS
4.1.x and Ultrix 4.x, for example.

I'm curious to know if you're saying that the above construct will no
longer be legal - it's certainly widely used in NetBSD sources, and is
pretty much the only way to declare things so that the code can
compile on both ANSI and K&R compilers.

			       _MelloN_