[Info-vax] problem with 64-bit pointers in C

[johnwallace4 at yahoo.co.uk]

On Thursday, 1 February 2018 14:07:00 UTC, Scott Dorsey  wrote:
> Stephen Hoffman  <seaohveh at hoffmanlabs.invalid> wrote:
> >
> >The existing backward-compatible 32-/64-bit environment is never going 
> >to be less than an utter and confusing and cryptic and arcane 
> >train-wreck.  Never.
> 
> This is true.  However, the same thing can be said of Windows and Linux too.
> 
> Solaris did it right.
> --scott
> 
> -- 
> "C'est un Nagra. C'est suisse, et tres, tres precis."

What exactly did Solaris do, and what it was called, and 
how it was invoked at build time (-taso or something?)?
I was trying to remember this a few days ago for some
reason, and failed. Other readers may never have seen
what Solaris did in this respect.

Obviously that was back in the days when source and 
behavioural compatibility with existing systems still 
had some value within organisations. 

Tru64 UNIX on Alpha, by comparison, had a largely 
pure64 environment. Did it help sales much (vs 
SPARC/Solaris)? Not in my experience. In fact lack
of compatibility with Solaris source probably 
significantly hindered Tru64 sales in the early days.

Then along came AMD64, which offered the kind of
compatibility with customers existing investments 
that some IT people nowadays think is not greatly 
relevant. 

But after compatible AMD64 arrived, IA64 lack of 
compatibility was one of a few factors that 
inevitably killed IA64. AMD64 wasn't much help to
SPARC/Solaris either.