[Coco] Glenside IDE + Multi-Pak + HDB-DOS

[Robert Gault]

Mark J. Blair wrote:
> I apologize if I seem pedantic, but we seem to be talking at right angles, Bill. I know how to jumper the Glenside IDE hardware. My question is specifically about Boisy's comment in the HDB-DOS source code from Toolshed which states:
>
> "note that the Glenside IDE controller does NOT work with an MPI using this current version of HDB-DOS"
>
> My question is:
>
> Is the addressing conflict at $FF7F the *ONLY* reason behind that comment in the HDB-DOS source code?
>
> I suppose I can also ask a related question: Why would a user without an MPI "probably want to use the 16-bit $FF7x driver"? Wouldn't the $FF5x configuration work in all cases?
>
> My motivation for asking these pedantic questions is that I'm designing a compact Glenside-IDE-compatible CF card adapter with ROM to fit in the lovely SuperPak cases, and I want to fully understand any compatibility issues so I can maximize its utility to anybody who might be interested in it. I'm left wondering whether there's a real need for it to support the $FF7x base address at all. Eliminating an unnecessary configuration switch would very slightly decrease cost and complexity.
>
>

Guys,

The issue with the Glenside controller probably is not caused by the MPI 
"reflecting" addresses at $FF7x. There looks to be some timing issue, maybe in 
the HDBDOS code. Boisy mentions READ problems not WRITE ones.

While this is not exactly the same as I have a Ken-Ton controller rather than 
the Glenside IDE, with my MPI system the hard drives are addressed in the $FF7x 
range with my Disto SCII controller set to the alternate $FF5x range.

Robert