[Oberon] Improving Oberon-0

[muller at inf.ethz.ch]

"Douglas G. Danforth" <danforth at greenwoodfarm.com> wrote:
> When trying to decide if one's machine is capable of running
> Native Oberon or Bluebottle such a Wizard written for MS Windows
> or Linux would seem appropriate.

I think time would be better spent improving the Oberon-0
environment, especially diagnostic information when things
go wrong.  There is no guarantee that the environment 
detected or presented by Linux or Windows is the same as 
what would be seen when Oberon is booted.

A big step forward would be to write code to switch back 
and forth between 32-bit protected mode and 16-bit real mode 
on-the-fly.  This would allow more BIOS functionality to be 
used.  E.g. if Oberon-0 runs into difficulties, it could use 
BIOS INT 10H to switch back to text mode and at least output 
a readable TRAP.

Extended BIOS facilities would also benefit normal Oberon, 
since it would allow using VESA 1.2 bank-switching in a 
generic display driver.

I've done some work in this direction with the V86 module,
which creates a virtual-86 mode task for executing 16-bit
BIOS code.  However, the environment is not perfect, since
some instructions need to be emulated in V86 mode (this 
has not been done).  A better approach might be to just
switch to plain old real mode.

If anyone is interested in attacking this problem I can
dig out some of my old test code as a starting point.

Another possible improvement for Oberon-0 would be to 
set up a mechanism for using the PCITools module to
scan for hardware.  This won't work on older machines
without PCI buses, but would be a big improvement on
newer machines.  A database of PCI ids and the matching
driver modules would have to be built.

-- Pieter

--
Pieter Muller, Computer Systems Institute, ETH Zurich / MCT Lab, Zurich
Native Oberon OS: http://www.oberon.ethz.ch/native/