[camlorn]

Well, yeah. or something like it. One of these days I'll read the official spec, but no one really needs to do that anymore (discounting for the moment the people who do). I mean, nowadays, the performance hit of using malloc/new is basically nothing.

But if you look back at diku and such, there's a lot of "hey, I know, let's grab a huge block of memory and implement our own because it's better/faster/whatever" going on. And then everything based off diku carries that forward, because no one wants to replace it. So yes, you have basically made my point for me, and I learned something new in the process.

But that code may be lying around, waiting to bite you when you least expect it. Be aware.