[Home] [Groups] - Message: [Prev in Group] [Next in Group]
8535: [MUD-Dev] Re: Bruce Sterling on Virtual Community goals
[Full Header] [Plain Text]
From: Chris Gray <cg@ami-cg.GraySage.Edmonton.AB.CA>
Newsgroups: nu.kanga.list.mud-dev
Date: Thu, 22 Oct 1998 22:05:26 -0600
Organization: Kanga.Nu
[Jon A. Lambert:]
>It also claims the operand stack is FIFO. I thought this was odd.
Definitely. If its FIFO it isn't a stack. By definition.
>Each method gets it's own stack frame. Perhaps this is peculiar
>to Microsoft's or Symantec's VM. The book is not really clear on
>whose implementation is being referenced. And I don't know if
>it's authoritative, it may well be misleading.
Could well be. The bit about the stack frame is correct, but when I
looked just now, I saw no mention of a frame pointer to point at it.
In the instruction format, locals are addressed via an offset within
the frame, so most implementations will likely use a frame pointer.
"Optop" sounds like the stack pointer, which they do define. Not
clear what "Vars" refers to. All non-local variables are inside classes.
I'd guess its the constant table.
--
Chris Gray cg@ami-cg.GraySage.Edmonton.AB.CA