Thanks, I'll give it a try. But as I wrote in the other posting: Usually all plugins work well with my conditions, I always try the original speed before reporting a bug. I never had any problems that was related to the speed of the emulation.

Alex

Hi Gerardo,

I know you are replicating the computers by using the original hardware. Please don't consider what I wrote as a self-experienced observation anything other than the chess software programs themselves (inside the ROMs). The fact that you are showing two ROMs for Regency possibly explains my experience with my Regency. It has occasional tendencies to move too fast or almost instantly which results in moves of possible lesser quality. Who knows there may be even third or fourth ROMs for Regency. You would know this better than I, as my interest is more about the chess programs other than being interested in how the different hardware influences the results of a program. The hardware is your domain.

In my tests I have only ever been able to 100% over 5 original tests replicate 226 positions with Legend and Concerto, every other one has a move deviance here and there. But even that would be difficult to state as a fact if there are multiple versions ROMs used with Legend and Concerto.

Therefore, just as you are doing it is important to separate and play on the computers under their original conditions (identify the different ROMs and separate them), unless you can 100% prove that two computers are 100% the same hardware and software. Because in my opinion one or the other would never = 100%. It requires both to be 100%. Your Mess approaches show this through your V1, 2, 3 versions, where many of them if not most provide different chess results. Otherwise, why bother with V1, 2 etc.

Based on what you are all doing with Mess I think you would be in agreement with this.

Best regards
Nick

Hi Hap,

It is funny how we always think the worst and react. If you play the chess programs you would know that their roots are very similar. Let's not rehash it all. If you read the history in Hiarcs you will know I was the biggest critic and had long battles with Ron Nelson, until I finally realized he was right. I was the one who first changed all reference to the name Ron Nelson on my Website and received a lot of critique for it in private as well. There is no program in the world (forget about chess, I am talking program) which does not have roots from somewhere. Programmers add ideas, remove them, change them, improve them... battle with hardware restrictions and so on, which ultimately results in a creation which is theirs. That applies to Ron Nelson and everyone else as well.

Best regards
Nick

Hello Nick,

I'm only a small part in all this, it's a teamwork of many people. I basically repair chess computers, reverse engineer and ROM dumping. But unlike many people, I actually play very little with chess computers these days.

I did an entry on the Shachcomputer.info Wikipedia explaining what the "ROM serial" is, a code that uniquely identifies the ROM of one of these Hitachi microcontrollers:

https://www.schach-computer.info/wik...tle=ROM_serial

But suppose for a moment two chess computers with exactly the same chess engine, the same opening book, basically identical hardware in terms of processing speed, RAM, etc. but one running with pressure squares and the other with magnetic sensors. The ROM dump will be very different, and yet they will play the same or very similar!

That is why it is also very important to study and compare the games... playing. All this is complementary, necessary and very interesting. One work of study does not exclude the other. The research done by hap, or by other members of this forum who dedicate time to the disassembly and study of the code stored in the ROM, sheds light, supports suspicions or sometimes disproves something that had been taken as valid until then.

Best,
Gerardo

Hi Gerardo,

Thanks, and yes, I fully agree with you, and I also respect and appreciate everyone in the Mess team and the incredibly hard work they put into it.

Best regards,
Nick

Previously I forgot to say that I don't do the hardware emulation. That's all done by hap, creating the so called MAME "drivers".

Best,
Gerardo

Hi Gerardo,

You wrote:

"But suppose for a moment two chess computers with exactly the same chess engine, the same opening book, basically identical hardware in terms of processing speed, RAM, etc. but one running with pressure squares and the other with magnetic sensors. The ROM dump will be very different, and yet they will play the same or very similar!"

It is the "very similar" which in this case I would be unsure about. Especially if it repeats itself occasionally in various tests that there is a move variation that occurs sometimes.

Are you able to read two ROM's through a program that highlights for example if it is binary were a 1101 was changed to 1011? The size overall would remain the exactly the same, but something had changed....even if it were by accident?

Or with ROMs and other chips are there not manufacturing tolerances (however minute) that could influence a slight deviation?

Asking as a layman, but with knowledge from Industry that everything has a tolerance for "pass" and "reject". Therefore, that means that even under "pass" there is a high and low "pass" that could influence some minor deviation on computer board A and same computer board B. Compound that by 20 different chips with (pass tolerances) = deviation might occur?

Layman's regards
Nick

Hi,

No, no, that doesn't happen. We are talking about binary data, zeros and ones. We compare the dumps bit by bit, there is no tolerance possible. If two dumps have for example one different bit, it can be: because the program is really different, due for example to a bug fix; because the data was stored in an EPROM and due to failure or aging that byte has been "corrupted"; etc.

But in the Hitachi microprocessors we are talking about, the ROM is stored in MaskROM and doesn't suffer "corruption" due to aging as it happens with EPROMs. So no, there is no tolerance... if 1 million of H8 microcontrollers are manufactured with a given binary code, and the programming of that ROM is verified in the manufacturing process, they will all contain the same data in the ROM unless there is a later damage...

Best,
Gerardo

Hi Gerardo,

I trust your methods 1000%. I am not worried about your Mess ROMs, my question is general around home chess computers. This is why I asked about board A and board B of a same computer and tolerances.

I don't think you can discount a possibility that deviations may occur even from the same production line.

Regards
Nick

PS... maybe I should add so what else could cause a move change if it is not the ROM. Speed? I know that people have built in other power supplies to their computers because of irregularities. Does the power distribution affect a move?

The human controlling the chess computer, not a joke.
The time duration spent on inputting the moves will not be the same everytime, allowing the chess computer to do more thinking(pondering?), or for example calculating a different random number that determines the next move.

The only time the hardware may cause a difference in playing style, hmm... On Fidelity chesscomputers maybe? They have a 555 timer IC for generating IRQs, that chip will not give identical timing for each chesscomputer.