> Actually, I'm not at all sure you are right about  Windows either.

Last I wrote code for Windows, several programs failed when you turn off
swap because you can specifically request storage (IIRC) from the kernel
that is "swappable" storage. You could have half a gig of RAM and have
programs fail if you turned off swap.

It's been a while since I've written anything on that level for win32, so
perhaps the semantics have changed, but I'd be leery of stating you can get
away without swap without actually trying it.

Anyone up for a ten minute trial?

Ray


Reply via email to