*** M E M O R U M ***

Hi all,

This is all very very interesting but way over my head.

Just my 2ps worth, but was there a standard used for CP/M and mice as that might help with deciding on a direction to go down.

A quick google did throw up https://www.seasip.info/Cpm/gsx.html

TTFN

Hi Paul,

The majority of CP/M software just used character displays. Customisation for a particular system just required defining things like the appropriate control codes used to, for example, position the cursor at coordinates x,y.

For the Memotech disk systems, the control codes are defined in the 80 Column card manual. Using these codes, software like NewWord or Wordstar could be configured for the system without much effort.

The graphics system extension that you referenced would need a custom extension to the Memotech CP/M BIOS and would not be an insignificant effort. Making use of it would require that the application software was written with the extension in mind and I am not sure how much CP/M software actually did so. There is certainly no support for it in any of the CP/M software typically used with Memotech systems.

So, as a I mentioned before, I don’t think that the effort of going down the software customisation route has any value. It would also require significant effort and I don’t think that anyone is likely to volounteer to do this work or be able to commit to completing it in a reasonable time frame - though I would be happy to be proved wrong!

Regards
Dave

Hi Dave,

Thanks for the reply. After I've ported x-windows to the MTX I'll be back in touch

Ta!

Or a GEM port ? There's source code available for that too......
http://www.deltasoft.com/downloads-gemw ... #GEMsource

I've finished upgrading the AVR software and the GAL to encode the mouse wheel movement and click.

Because of limited pins on the 16V8 GAL I'm only able to drive 2 sense lines. All 16 input pins on the GAL are used, 9 are needed for the input from the AVR and 7 for keyboard drive lines. Leaving just the 2 for output. Having chosen to implement the cursor keys and home, plus enter. The additional 3 mappings need to be on KB6 with enter. The middle button click as insert, and the mouse wheel as [ and ] not so much use in existing software, but the potential is there.

Moving to a 22V10 would allow all 8 drive lines to be monitored, and an extra 3 sense lines, increasing the key options. Having the when "press" the control key combinations that Newword uses for page up/page down.

Having worked through the problems caused by a failing GAL, I've finally been able to test the mouse on more than just Reversi and MTX basic.

Since I've not got Supercalc installed on any of the drives, the most important test is Newword, and the result of the test was .....

Disappointing

Very disappointing

The sort of mouse motion that would move the pointer on this PC across the full with of the screen, moves the cursor in newword two or three characters at the most.

Getting the cursor to the right side of the screen was a joke. The keyboard scanning routine is just too slow to be practical.

What if the CP/M BIOS set up a frequent interrupt to read the mouse keys?

Tony Brewer wrote:What if the CP/M BIOS set up a frequent interrupt to read the mouse keys?

The problem then becomes, what's going to use it.

Interfacing the mouse via the cursor keys was done to try and make it fit in with existing software.

As Bill pointed out, there's much better ways of interfacing the mouse if you're willing to write the software to handle it. After all, I'm converting the 4 bytes that the mouse send to the microcontroller into just 9 bits of keyboard info. To use the mouse to it's best effect, you'd want all 4 bytes of data available to the CPU.

Having the AVR fill a pair of 74HCT670 4 bit register files for the CPU to read back would do that. The existing size of GAL should be able to cope with the CPU side interfacing.

Somehow I can't see anyone doing a new mouse operated wordprocessor (or any other application) for CPM at this point in time.

A version of the MTX videowall could be driven by a mouse, but that was via serial port.