"self" vs other names

[Ignacio Vazquez-Abrams]

On 3 Oct 2001, Samuel Schulenburg wrote:

> grante at visi.com (Grant Edwards) wrote in message news:<slrn9rktbp.rfk.grante at isis.visi.com>...
> > In article <mailman.1002072003.6214.python-list at python.org>, Ignacio Vazquez-Abrams wrote:
> > >On Wed, 3 Oct 2001, Tim Hammerquist wrote:
> > >
> > >> What would the world have been like if CP/M had won over (PC|MS)-DOS?
> > >
> > >Not that much different; AFAIK MS-DOS was based on some of the principles of
> > >CP/M (8.3 filename, etc.).
> >
> > PC-DOS was an out-and-out copy of CP/M.  The layout of FCBs was the same,
> > the methods to call BIOS were the same, the layout of executable file
> > headers was the same...
>
> As I recall CPM was more flexable with memory allocation than DOS.
> It's been a long time ( Twentyone years),but CPM used a jump table to
> access its boot sequence. DOS had that dumb 640 K limit.
> Remember BIll Gates quote "640K is more memory than anyone needs", or
> somthing like that.

Although Bill Gates did say something along that line, the 640k limit was
actually created by Intel and IBM.

Intel in their great forward-thinkingness decided that they wanted the 8086 to
be compatible with the Z80 (yes, your shiny P4 is a Z80 ripoff!), and so
instead of starting off with a 32-bit address space they implemented a
horrific 16-bit segment/16-bit offset addressing scheme which flattened out to
a 20-bit address space, giving 1 megabyte of memory space.

IBM, new in the PC business, decided that their computer would reserve the top
6 segments (A000-F000, or 384k) for RAM and ROM situated on interface cards.
They figured that it was enough at the time; people had barely gotten used to
having 128k in their machines.

-- 
Ignacio Vazquez-Abrams  <ignacio at openservices.net>