Re: RFC 47 (v1) Universal Asynchronous I/O

[Bart Lateur]

On Tue, 8 Aug 2000 22:49:13 -0400, John Tobey wrote:

>As has been mentioned on a nearby subthread, we're probably just going
>to store the function call stack, not top-level variables or code.

That doesn't sound useful, at all.

For example:

        sub foo {
           my($x, $y) = @_;
           bar(2*$x, $x+$y);
           print $x, $y;
        }
        sub bar {
            ...
            continuation; # stop and restore to this point;
            ...
        }

When bar returns, the values of $x and $y will simply be lost.

I can live with files that all have been closed, database connections
that have been lost, etc. But *all* variables reset, global and lexical?
You'll get a pretty useless system, then.

-- 
        Bart.