Re: IIgs RAM drive question

[dempson@actrix.gen.nz (David Empson)]

Jeff Blakeney <CUTblakeney@home.com> wrote:

> It is a built in feature of the IIgs.  No modifications need to be
> made to the IIgs hardware to allow this to happen so it is, by
> definition, built in.
> 
> Admitedly, the designers made not have done this intentionally but
> that is how it turned out.

I'm certain it was intentional.  The SCC has to support two serial
ports, so it is not reasonable for it to appear in the DEVSEL address
space of either slot 1 or 2, unless it is duplicated in both of them.

In addition, AppleTalk needs to be able to access the SCC without either
slot being set to the built-in device, so it makes more sense to move it
out of the slot-specific address space.

I don't think Apple ever intended to let the serial hardware be used
directly by application code, except in limited ways under control of
the serial firmware.  The capabilities provided by the IIgs serial
firmware are a vast improvement over anything available on earlier
machines.  The main problem is that the interface to the firmware
(Pascal 1.1 driver in the slot firmware) is a major bottleneck,
especially when dealing with high speeds.

> >It has nothing to do with the firmware, just the hardware.
> 
> Well, I kept saying "hardware and firmware" because they are very
> closely tied together and the way the firmware works when the slot is
> changed between the "Modem Port" and "Your Card" is relevant.

The accessible parts of the serial firmware are completely switched out
if the slot is set to "Your Card", so its behaviour isn't particularly
relevant.

There is still the interrupt handler, but that is also removed from the
picture when the application's driver replaces the vector.

> However, you seem to be agreeing with me that it is mostly the
> hardware design that allows you to access the serial port hardware
> directly even when the firmware has been told that slot 2 is set to
> "Your Card".  This still says to me that Adalbert was right that it is
> the hardware that allows you to do this rather than any particular
> piece of software.

Well, he said it was a feature of the "ROM 01", which is an
understatement.  It is a feature of the IIgs (all versions), if
anything.

I still can't see what the point is describing this as a feature of the
IIgs, since you could say the same about every feature provided by every
application on the machine - no software could do what it did if the
hardware didn't allow it.

I'm trying to distinguish between "normal" capabilities (e.g. using the
serial firmware) and "exceptional" capabilities (e.g. using the serial
hardware directly).

> Saying that ProTERM is providing that ability is kind of like saying
> that MUG! allows you to access MFM encoded disks on hardware that
> isn't capable of reading MFM encoded disks.

No it isn't.  ProTerm is providing a feature which is not available
under normal circumstances, but is possible given the hardware features
of the IIgs.

A similar example would be the sound input hardware provided by the
Ensoniq.  You need special software to make use of this feature, since
there is nothing significant in the firmware or toolbox to support it.

Another example is the vertical and horizontal scan counters (and their
use in tricks like displaying text in the border area, and implementing
3200 colour mode).  The hardware may allow these sort of tricks to be
pulled, but it wasn't necessarily an intended feature of the machine.
The software has pulled a rabbit out of the hat.


In any case, this seems a rather pointless argument.  We are agreeing
that the IIgs is great.  I just prefer to credit the software with
squeezing exceptional capabilities from the hardware.

-- 
David Empson
dempson@actrix.gen.nz
Snail mail: P O Box 27-103, Wellington, New Zealand