Re: How much RAM can debian support?
On Sun, 11 Oct 2009 08:10:11 -0700
Andrew Sackville-West <andrew@farwestbilliards.com> wrote:
> On Sat, Oct 10, 2009 at 08:23:10PM -0700, Marc Shapiro wrote:
> > Andrew Sackville-West wrote:
> >> On Sat, Oct 10, 2009 at 12:50:41PM +0200, Magnus Pedersen wrote:
> >>> Alex Samad wrote:
> >>>
> >> [...]
> >>>> God can you remember when 16M used to be a lot ...............
> >>>>
> >>> My first one, a commedore vic-20, I think I was 8 years old and it
> >>> was my first real love...
> >>> 1MHz CPU I think is was a 6502
> >>> 5K RAM
> >>
> >> but IIRC, you could only access 3.5 as the other 1.5 was taken up by
> >> the system. And I could outtype my modem (300 baud) with one hand.
> >
> > I remember my first 300 baud modem.
>
> what was really fun was the upgrade to 1200 baud. Wow that thing was
> fast!
>
> > And downloading e-mail from
> > Compuserve with a script that would connect, download and disconnect so
> > as not to spend any more time actually connected (at per minute charges)
> > than was absolutely necessary.
> >
> > Maybe those weren't the "Good, old days."
>
> yes and no. no for a lot of really obvious reasons, but yes because
> you had complete control and *understanding* of the machine. That's
> what makes it so great in my memory. I could really understand what
> the whole thing was doing, and thus manipulate it without to much
> difficulty. Don't get me wrong though, there were (and still are)
> plenty of folks who could make those machines, particularly the C-64,
> really jump through some crazy hoops in ways I could never imagine.
>
There were times when software would modify itself in memory due to lack of
memory space. I also saw contests where people wrote programs that would read
the same start to end and end to start (forgot what that is called). I've got
copies of papers of how to efficiently transpose a matrix when only a couple of
lines fit in memory (and we are talking 512x512 8bits per pixel)
On the other hand hand even on more modern hardware I know people who would
program debugger breakpoints into the code and worse.
> A
Reply to: