Q-bus to CF [was: IOmega]

[Sridhar Ayengar]

Jerome H. Fine wrote:
> I am not sure if this suggestion is easier or even useful.
> 
>> From what I gather of looking at the MSCP protocol from
> the DU(X).SYS MSCP device driver in RT-11, there may
> be an easier implementation that is just as useful.
> 
> John Wilson set up an interface for an HD(X).SYS device
> driver under Ersatz-11 in order to provide variable
> sized devices until the MSCP emulation had been coded.
> 
> While John's version of HD: needs a bit of modification
> to work well with mapped monitors, it is still only for
> RT-11.  On the other hand, for the past 6 months, I have
> been adding additional features which include being able
> to handle full 22 bit Qbus / Unibus interface for mapped
> monitors.
> 
> The real HD: advantage is the simplicity of the interface.
> While there are 8 I/O registers, setting them up when
> interrupts are not used takes less than 20 PDP-11 instructions
> including the 22 bit address.  If you look at the hobby
> version of E11, John includes the code for HD(X).SYS for
> RT-11, although as I mentioned some changes are needed for
> mapped monitors when the user buffer is above 1/4 MByte.
> 
> If anyone ever produces a Qbus / IDE (or any other PC
> current drive) controller, I will gladly enhance the HD(X).SYS
> device driver for RT-11 to use a DU(X).SYS translation
> type table.

Wouldn't the concern then be that it would only be supported in RT-11, 
and/or it would not be bootable?  One of the primary motivations for me 
in this exercise would be CD-ROM booting.  I'm seriously considering an 
off-the-shelf solution for the ATA/ATAPI side to save myself some 
effort.  No sense reinventing the wheel.

Peace...  Sridhar