Re: Hard Sci-fi?

Wayne Throop wrote:

> Well.  There's a miniscule fraction on which it's possible, right?
> So, one would expect that it's possible to configure computers to
> allow compilation.  It had better be possible to so configure a
> computer, otherwise, you'd have to hand-assemble the instruction
> store, and the temptation to make it an interpreter would be
> overwhelming.  I mean... look at what went into Apple ][ roms.
> An interpreter.  Early on, the ram *was* mostly used only for data.
> Typically, the data bas an integer basic program.

Yeah.  Michael summed it up quite well; the exploits we're having are 
general software problems, not something that can be solved by switching 
to a different architecture.  The fundamental point here is that 
programs are devices which turn data into action, no matter how trivial 
that data may be.  The problem is that bad data can cause the program to 
function inappropriately if its programmers were not sufficiently careful.

Or, to put it another way, if you somehow forced all programmers to use 
a Harvard-style architecture, you'd end up with interpreters doing 
everything anyway, which would just force the problem down to the 
interpreter level, not to the protected code area.  And which is largely 
happening on its own for plenty of other reasons, quite frankly.

-- 
Erik Max Francis && max@[EMAIL PROTECTED]
 && http://www.alcyone.com/max/
  San Jose, CA, USA && 37 18 N 121 57 W && AIM, Y!M erikmaxfrancis
   Woman was God's _second_ mistake.
    -- Friedrich Nietzsche