MG Explorer for the Geneve-any interest?

[+FALCOR4]

First, is there already a version of the MG Explorer for the Geneve out there? We haven't been able to find one yet. Is there any interest?

 

Years ago Mike Dodd wanted to modify it to work on the Geneve. He spent some time recoding and returned a disk with his initial work. I have nothing after that and I don't know Mike's whereabouts/status today.

 

So, if there is interest and if someone would like to take up the task I would be happy to post what I have. I haven't taken an in-depth look at the changes and I probably won't be able to engage too much on this project given time constraints. If you've good with coding, have good knowledge of the Geneve, have the time and want to take this on let me know. Here is a synopsis of the notes page that Mike sentat the time:

 

NOTES ON GENEVE VERSION OF MG EXPLORER
Modified for the MYARC Geneve 9640 Computer by Mike Dodd

Nature of this release of Explorer:

This version will probably work on a 99/4A, but
slllowwwllllyyyy, due to all CRU keyscans being removed.. It would be much
better to use Bytemaster's version of EXP for the /4A - this version was
meant for the Geneve..

There are two versions on the disk: EXP and EXP1.. EXP loads at
>B9C4 and is designed for operation from the Editor/Assembler
environment.. EXP1 loads at >A000 and is designed for the Extended BASIC
environment.. It can be loaded with the LEXP utility provided on the
Bytemaster release disk.. Both versions are >462C bytes long..

Changes:
[1] EXP used CRU keyscan to check for CTRL 2-5 when in continuous
execution mode.. Changed to use normal keyscan link..
[2] EXP used CRU keyscan in memory editor to check for FCTN-SHIFT
cursor movement.. Removed..
[3] EXP used CRU keyscan to scan for SHIFT when in continuous
execution mode, for purpose of placing in Turbo mode.. Removed..
[4] EXP had many "Anti-Piracy" checks, to prevent changing the load
interrupt vectors, and looking at/altering EXP's memory.. These have been
removed..
[5] All auto-repeat code has been removed - they were too fast on
the Geneve's higher speed, and the Geneve's IBM keyboard has auto-repeat
built in..
[6] If the PC reaches >02B2, indicating a keyscan, the Geneve will
execute the keyscan at high speed and return to *R11.. This was changed as
the Geneve's keyscan involves memory paging, and would page EXP out of
memory, resulting in a lockup.. Unfortunately, this still does not work
perfectly, and I'm not sure how to fix it.. The problem is this: when you
STEP an instruction, either in single step mode or continous mode, EXP
scans the keyboard, using the standard link (see [1] above).. This keyscan
almost always grabs a new key first, before the AP program gets a chance
to read the keyboard.. Thus, when the AP program scans the keyboard, the
keyscan returns a status of >00, since, as far as it's concerned, that
key was already pressed.. Which is true.. Then, of course, the AP program
scans the status byte and sees that nothing has been pressed, and scans
again.. This can continue on forever.. I honestly don't knoww what the
solution to this is.. I am welcome to any ideas..
[7] EXP destroyed all memory when CTRL = was pressed to exit.. This
has been removed, so that on CTRL =, EXP executes a BLWP @>0000..
[8] colors changed to suit my perverted tastes (I have a monochrome
monitor).. You don't like them? Tough - wait for the final release of
this.. I'll try to have easily changable colors in it..

Changes planned:
[1] Would like to modify EXP to have option to execute DSR calls at
full speed, so that programs may call the DSR and still work on the EXP..
This shouldn't be too difficult..
[2] Would like to modify EXP to reside in different memory pages
than AP program, so that the AP program could take up as much memory as
it wanted.. However, this may be next to impossible to do.. I'll have to do
a detailed study of EXP's usage of AP memory.. If the only users of it are
STEP and DASSEM, it might not be too difficult.. Otherwise........ could be an
incredibly neat trick.. Sure would be nice though - would make EXP the
ultimate debugger..
Keep in mind that unless change [2]
above is accomplished, memory is very tight.. For that reason, such large
features as, for instance, disk catalogs, printer dump, help screens,
etc.. will not be considered, _unless_ I can get the above change made..
Even then, I will want to keep it to 32K maximum (it's about 17..5K now,
so there would definately be room for expansion), counting code and
buffers.. BUT, I would not bet great amounts of money on the success of
changing the EXP to reside in different memory banks, so keep it
reasonable..

Edited May 28, 2018 by FALCOR4

[Opry99er]

I don't have a Geneve, but I think this would be an excellent addition to the "archives".

 

There may not be anyone immediately wanting to take this on, but perhaps some day....

[+FALCOR4]

Agreed. I'll put my CC FDC in maybe this weekend and check the disk to make sure I captured all the material. Am using the Myarc FDC now and it's only a 16 sector/track on the DSDD but the disk from Mike is 18 per. Shouldn't make a difference just reading the disk but stranger things have happened......

[+InsaneMultitasker]

Interesting... I know I've seen the above readme information before. I wonder if this program got out into a few hands long, long ago? Can't say that I recall running the program on my Geneve though back then I was busy with programming my own software.

[+9640News]

The source code would be interesting to see and maybe there is something to be learned. However, I am not sure how likely someone would be to use it. I only say that because with the debugger for MAME, it is far more powerful than I think Explorer is.

 

Beery

[+retroclouds]

The source code would be interesting to see and maybe there is something to be learned. However, I am not sure how likely someone would be to use it. I only say that because with the debugger for MAME, it is far more powerful than I think Explorer is.

 

Beery

Agreed, but if you think about debugging assembly on the TI-99/4a itself, it is more than impressive what they pulled of there. Do not have a Geneve, but would be interesting to see if explorer could be reassembled to run from the cartridge space. Combine that with a good programmers editor, an assembler running from the cartridge space and you have a heck of a programmers environment. That being said I learned TMS9900 assembly language on the PC with classic99. But I can imagine using the TI as my development box for retro programming.

[+acadiel]

Falcor, was any of this on the hard drive image we released already?

[+FALCOR4]

Falcor, was any of this on the hard drive image we released already?

 

No, it wasn't in the original files I pulled off the HD. I'm still finding all kinds of stuff from back then; I ran into this one on one of the floppies that Mike sent. I'll post it as soon as I know that I have it all. I was finally successful in getting everything off the drive, I sent you an email. GoTek to the rescue! Much thanks for that.

Edited June 3, 2018 by FALCOR4

[+FALCOR4]

 

No, it wasn't in the original files I pulled off the HD. I'm still finding all kinds of stuff from back then; I ran into this one on one of the floppies that Mike sent. I'll post it as soon as I know that I have it all. I was finally successful in getting everything off the drive, I sent you an email. GoTek to the rescue! Much thanks for that.

 

Acadiel, you have them now. I took a cursory look at the source files and didn't see anything wonky going on so I'm going to assume they are intact.

[+FALCOR4]

Looks like these files are good. Posting..... If you're out there, Mike Dodd, would love to talk to you

 

DSKA0028.hfeDSKA0029.hfeDSKA0030.hfe