Novag Micro II & Clones by Kaplan, not Kittinger!

[bataais]

For previous investigations see also: https://www.schachcomputer.info/forum/s ... php?t=5760

I once again devoted myself to the question of whether the Novag Mirco II and clones (Micro III, Presto, Allegro, Piccolo, Octo) could be programmed by Julio Kaplan, as Jan Louwman claimed in Computerschaak, Feb. 1983.

I found an interview with Julio Kaplan in CSS 1993, 4th edition, page 9 (Socrates' apprenticeship and wandering years) where he says:

"Other manufacturers, including Mephisto and Novag, used Heuristic-Software programs as well."

The Mephisto is probably the Monaco, a clone of the Express 16K.
Kaplan's Novag should be an early model, either Micro Chess (F8) or Mirco II, which has an Intel MC (8049, 2k ROM, 128Bytes RAM)t, as well as the Mattel (8050, 4k ROM, 256Bytes RAM) , Kaplan's first unit.

When comparing the disassemblies of Mattel and Mirco II, I found an identical routine that sets up the chessboard in the basic position:

Mattel Computer Chess:

Code: Select all

1:0BA: B8 D9 mov  r0,#$D9
1:0BC: B9 80 mov  r1,#$80
1:0BE: F8    mov  a,r0
1:0BF: A3    movp a,@a
1:0C0: C6 D8 jz   $0D8
1:0C2: A1    mov  @r1,a
1:0C3: 19    inc  r1
1:0C4: B1 10 mov  @r1,#$10
1:0C6: 19    inc  r1
1:0C7: 19    inc  r1
1:0C8: 19    inc  r1
1:0C9: 19    inc  r1
1:0CA: 19    inc  r1
1:0CB: B1 18 mov  @r1,#$18
1:0CD: 19    inc  r1
1:0CE: 43 08 orl  a,#$08
1:0D0: A1    mov  @r1,a
1:0D1: F9    mov  a,r1
1:0D2: 03 09 add  a,#$09
1:0D4: A9    mov  r1,a
1:0D5: 18    inc  r0
1:0D6: 04 BE jmp  $0BE

1:0D9: 02 04 01 03 05 01 04 02

Mattel board in RAM:

Code: Select all

02 10 00 00 00 00 18 0A
04 10 00 00 00 00 18 0C
01 10 00 00 00 00 18 09
03 10 00 00 00 00 18 0B
05 10 00 00 00 00 18 0D
01 10 00 00 00 00 18 09
04 10 00 00 00 00 18 0C
02 10 00 00 00 00 18 0A

Novag Mirco II:

Code: Select all

0:008: B8 23 mov  r0,#$23
0:00A: B9 40 mov  r1,#$40
0:00C: F8    mov  a,r0
0:00D: A3    movp a,@a
0:00E: C6 2C jz   $02C
0:010: A1    mov  @r1,a
0:011: 19    inc  r1
0:012: B1 01 mov  @r1,#$01
0:014: 19    inc  r1
0:015: 19    inc  r1
0:016: 19    inc  r1
0:017: 19    inc  r1
0:018: 19    inc  r1
0:019: B1 09 mov  @r1,#$09
0:01B: 19    inc  r1
0:01C: 43 08 orl  a,#$08
0:01E: A1    mov  @r1,a
0:01F: 19    inc  r1
0:020: 18    inc  r0
0:021: 04 0C jmp  $00C

0:023: 04 03 02 06 05 02 03 04

Micro II board im RAM:

Code: Select all

04 01 00 00 00 00 09 0C
03 01 00 00 00 00 09 0B
02 01 00 00 00 00 09 0A
06 01 00 00 00 00 09 0E
05 01 00 00 00 00 09 0D
02 01 00 00 00 00 09 0A
03 01 00 00 00 00 09 0B
04 01 00 00 00 00 09 0C

So Louwman was right after all. The Novag Mirco II is actually a Julio Kaplan program, a slimmed-down version of Mattel Computer Chess!

Disassembled Regards,
bataais

[Berger]

Hello bataais,

Wow! Great research work indeed. And the conclusions drawn are very interesting.

Best,
Gerardo

[Larry]

Hello bataais,

Wow! Great research work indeed. And the conclusions drawn are very interesting.

Best,
Gerardo

Yes a big thanks for this research. I did always wonder why the
Novag micros were pitifully weak, since they were listed
over on .info as authored by Kittinger, a strong programmer.
L

[Tibono2]

Thanks bataais, great work.

I trust the code... and your conclusion.
I updated the Novag Octo authorship on my website.

Cheers,
Eric

[bataais]

Larry, I think it's not about who is the better programmer.
With so little RAM and ROM, it's a wonder, the little machines can play by the rules at all.
Dave Kittinger even wrote a smaller program in 2K ROM and 96Bytes of RAM! (Prelude/Solo/Escort). It is playing in the same league as Novag Micro II.

None of the 2-4K programs go beyond 1350 ELO, chess is just too complex for so little hardware resources.

[Mike Watters]

Many thanks for your research bataais.

I think there may be many more surprises out there to be discovered.

[Berger]

Hello,

I think there may be many more surprises out there to be discovered.

I just bought a Novag Allegro 4. I hope to receive it next week. As soon as I get it, I'll dump the rom and send it to bataais for analysis, and to hap for emulation on MAME.

Best,
Gerardo

[bataais]

Hello berger,

the Allegro 4 has a Toshiba TMP47C400 4bit MCU, I don't know if there is an easy method to dump it. Good luck! This will certainly be a Kittinger

But there is still another 8049 chess computer, not already dumped:
The Fidelity Eldorado, by Ron Nelson, anno 1990. I've got one, it plays better than Mini Sensory II.

2K regards,
Michael

[Berger]

the Allegro 4 has a Toshiba TMP47C400 4bit MCU, I don't know if there is an easy method to dump it. Good luck! This will certainly be a Kittinger

OMG... Money wasted!

It is not a kind of computer that I usually buy except for reverse engineering...

Best,
Gerardo