Ron Nelson Ever Copied, Used , Cloned the Spracklen?

This forum is for general discussions and questions, including Collectors Corner and anything to do with Computer chess.

Moderators: Harvey Williamson, Steve B, Watchman

Forum rules
This textbox is used to restore diagrams posted with the fen tag before the upgrade.
Post Reply
User avatar
spacious_mind
Senior Member
Posts: 3999
Joined: Wed Aug 01, 2007 10:20 pm
Location: Alabama
Contact:

Post by spacious_mind »

Monsieur Plastique wrote:To do so is I think a disservice. Furthermore, without the necessary qualifications to authorship, there is no real explanation for the drop in strength in later machines, since as I say it cannot be explained purely by a slower processor and lack of pondering.
You are right, especially around quality I have about a dozen LCD's that stopped working on the LCD Chess computers after 1 or 2 battery changes. I am still looking to find another New York Times due to this. Hard to find. 1 Battery change and the LCD started removing chess pieces from the board. I have not been able to play it since. I even bought another Chess Station in case the Pocket Chess does the same.

Ron however did explain that his interest was the mass market therefore hard to argue against that when it comes to loss in play strength as times progressed on new models.

Best regards
Nick
User avatar
Steve B
Site Admin
Posts: 10140
Joined: Sun Jul 29, 2007 10:02 am
Location: New York City USofA
Contact:

Post by Steve B »

Monsieur Plastique wrote:
I am not really the Nelson fanboy that many others are.
nice personsal attack Jon
i think its hilarious that those who turned out to be dead wrong about Nelson and the GM ..now go on the offensive and are attacking and lecturing those that turned out to be right about Nelson and The GM

your comment is both rude and offensive and if it were directed at a specific individual(other then me)then i would delete it

i wont mention the names they are calling me on another forum ..so i guess this form of poor behavior is catching on

Steve
User avatar
Monsieur Plastique
Senior Member
Posts: 1014
Joined: Thu Jul 03, 2008 9:53 am
Location: On top of a hill in eastern Australia

Post by Monsieur Plastique »

spacious_mind wrote:I think if you go back and read all of Nelson's comments, he explained that for Alexandra a lower quality chip was used to the H8, which resulted due to not enough ROM space the removal of the Attack Tables as well as the ability to Ponder to make it all fit.
If you read all of my previous post you will see that I mentioned and acknowledged all of those factors, however I still have difficulty believing that these RAM and CPU hardware handicaps are largely responsible for such a massive strength difference between the GM and later products. I also pointed out that in my additional experiments with the Alexandra program, I did not find any subjective improvement with extraordinary long time controls (as high as 40 in 3 hours versus 40 in 90 which is effectively a doubling of MIPS).

Without knowing the exact relative MIPS capabilities of the differing hardware (GM versus Alexandra) it is of course difficult to come to a quantifiable judgement, however at this juncture I find it extremely difficult to believe that the attack tables did not count for a substantial component of the strength difference, since the hardware and pondering are only worth so much.

Perhaps we could even intelligently speculate if we knew the exact MIPS and worked on the paradigm that a doubling in MIPS on a non-hash dedicated machine is worth 60 points. Pondering on a non-hash machine is obviously only useful if you get a ponder hit and since this only occurs typically for a small portion of moves, I would be reluctant to say this would add more than, say 30 points at the most.

So if we take the current strength difference at 40 in 2 at Schachcomputer as 300 points, we have to take the hardware and pondering differences off that ELO difference and arguably we are then left with the benefits conferred by the attack tables and the "Kaufmann Consultancy".

I think it would be an interesting experiment. Even a theoretical result based an accurate knowledge of the relative MIPS would - at least from my perspective - give me a fair idea of how important these factors are as they specifically apply to the GM / Mirage / H8 Igor versus everything else that came afterwards.
Chess is like painting the Mona Lisa whilst walking through a minefield.
User avatar
spacious_mind
Senior Member
Posts: 3999
Joined: Wed Aug 01, 2007 10:20 pm
Location: Alabama
Contact:

Post by spacious_mind »

Monsieur Plastique wrote:
spacious_mind wrote:I think if you go back and read all of Nelson's comments, he explained that for Alexandra a lower quality chip was used to the H8, which resulted due to not enough ROM space the removal of the Attack Tables as well as the ability to Ponder to make it all fit.
If you read all of my previous post you will see that I mentioned and acknowledged all of those factors, however I still have difficulty believing that these RAM and CPU hardware handicaps are largely responsible for such a massive strength difference between the GM and later products. I also pointed out that in my additional experiments with the Alexandra program, I did not find any subjective improvement with extraordinary long time controls (as high as 40 in 3 hours versus 40 in 90 which is effectively a doubling of MIPS).

Without knowing the exact relative MIPS capabilities of the differing hardware (GM versus Alexandra) it is of course difficult to come to a quantifiable judgement, however at this juncture I find it extremely difficult to believe that the attack tables did not count for a substantial component of the strength difference, since the hardware and pondering are only worth so much.

Perhaps we could even intelligently speculate if we knew the exact MIPS and worked on the paradigm that a doubling in MIPS on a non-hash dedicated machine is worth 60 points. Pondering on a non-hash machine is obviously only useful if you get a ponder hit and since this only occurs typically for a small portion of moves, I would be reluctant to say this would add more than, say 30 points at the most.

So if we take the current strength difference at 40 in 2 at Schachcomputer as 300 points, we have to take the hardware and pondering differences off that ELO difference and arguably we are then left with the benefits conferred by the attack tables and the "Kaufmann Consultancy".

I think it would be an interesting experiment. Even a theoretical result based an accurate knowledge of the relative MIPS would - at least from my perspective - give me a fair idea of how important these factors are as they specifically apply to the GM / Mirage / H8 Igor versus everything else that came afterwards.
Ok Now I am starting to follow you. I had started my tests a while back but never completed them for Alexandra.

Here is test game 1:

Image

It scored very well in Test Game 1 outscoring Igor and Mirage that scored identically but not outscoring Ivan.

Here is test game 2:

Image

Test Game 2 was atrocious. One of the worst scores ever recorded. Even Nelson's Peri Beta and Silver Bullet scored better. It was so bad that I repeated the test twice in case I was dreaming but the second test actually came out worse than the first test.

I will need to complete the tests and get its rating soon. As well as the other mentioned Excaliburs of its class.

So I am not necessarily disagreeing with your conclusion regarding strength.

Best regards
Nick
ChessChallenger
Member
Posts: 50
Joined: Mon Dec 21, 2015 7:48 pm

Re: Ron Nelson

Post by ChessChallenger »

Hi,
I have returned from my travels to spend Christmas with my family.

I have not enjoyed reading most of the postings in my absence, but some were very meaningful and heartfelt.

First I want to set something right in a post I made, I said I wrote and desktop published all my Instruction Manuals. I think it was a case of last in first out memory. For you engineers, are there any on here, LIFO and not FIFO buffer.

When I started at Excalibur, Al Lawrence had his own company that consulted on writing Instruction Manuals for Excalibur. He wrote the first manuals for my chess products. They were wonderfully written and typeset, using QuarkExpress. I asked and received all content files he created.
When I went on my Sabbatical and returned, it was about the time we were moving into the famous Rook Tower building with the Chess Museum.
Excalibur needed MORE product than one (me) person could design and program. So Al Lawrence was brought on, as VP of Product Development, to work with our current HK Factories and new Factories Shane & Mike found had interesting products.
NO computer Chess products.
He could no longer do my manuals with as much attention, so I bought the QuarkExpress software and did all my manuals from then on.
Al Lawrence is another brilliant person, Fidelity and Excalibur had as friend, worker and advisor. At one point when it was asked of him, he lived in Hong Kong running our HK office for 6 to 9 months. My fondest memory of Al is seeing him play a simultaneous chess exhibition in downtown South Miami near where I lived.
ChessChallenger
Member
Posts: 50
Joined: Mon Dec 21, 2015 7:48 pm

Re: Ron Nelson

Post by ChessChallenger »

So I corrected a wrong, with Al Lawrence. By the way it has been 40 years since I invented Chess Challenger on my Altair 8800 home built computer. And Excalibur did go bankrupt and it's assets were bought by EB Brands, so again, all was not rosy towards the end. So things are coming back, perhaps with memory errors from age and some from pain of recalling.

Fidelity Designer 2265, my most fondest Fidelity memory, a crowning achievement at Fidelity. A certified consumer computer chess master.
But what came before, the Mach IV...or some name like like that.
The Spracklens were maxed out.
Kathy, in charge of evaluation functions with Chess Masters in her office advising her, was looking for some other blue sky approach. Dan was now in complete charge of 68000 program advancement. He had executed on hash tables perfectly, so end game was fantastic. I gave him the multiprocessor hardware to use two 68000 programs (on my chess cards) with medium coupling, after I had developed my twenty one 6502 multiprocessor machine I entered at the ACM with loose coupling. But it was not enough, and costly.
All rested on null move search technology. He could not make it work to give meaningful improvement.
They were in California and I was in Miami, but I had something, I had made.
An auto-tester I built, using a competitors product playing a 68000 chess card, I had designed, with the Fidelity Spracklen engine.

I could have many automated matches, printing out PRV and scores. But it was my watching the games and the PRVs like you do, and we did at ACM tournaments that started me asking questions of Dan. Why can we not generate checks in the quiescent search? He said because it would blow up the search and slow down. Ok, I said, but what if we only generated checks that didn't occur as often, like a knight check that forked a major piece. He said, ummm,,, that would not take much and the search would not blow up.
So that is how we slowly started developing a tactical quiescent search that had all of the things a strong chess player explores when thinking of a tactical position. But I would see that the PRV was missing these obvious strong player "tricks" and have Dan look to see if he could add them.

Because of the attack map, all this type of information was easily divined. At the Micro Tournament in Spain, it was music to my ears to have the Chess Master commentator, perhaps Mike Valvo, say The Fidelity unit was playing moves it had never before been capable of playing. Just like the Masters we played to get the certified rating, who were amazed.

I used this same type of tactical threat generation on the H8 machine, since I had attack maps with the needed information.
ChessChallenger
Member
Posts: 50
Joined: Mon Dec 21, 2015 7:48 pm

Ron Nelson & Larry Kaufman

Post by ChessChallenger »

Even now I have to catch myself when thinking why didn't I bring in Larry Kaufman to help in Fidelity's computer chess development.
Sid Samole gave the Spracklen's in their California office everything they asked for. They sequentially went through at least two or three Grand Master advisors.

And Larry was in the Miami area and would come in to see me or Sid and we would talk. But never meaningful to me.
We talked about his Computer Chess Reports Mag & results. He played many game of consumer computer vs. consumer computer chess like you do, and report the results, and some general opinions.
But I think it was the years of watching these computer vs computer matches that his Chess Master skills began to see why they were making the moves they were making. He could start thinking like the programs, and so he understood the programs. Not just the evaluation function, but the tactical search, and selective search that Lang used in certain positions.

So when he came to me at Excalibur, he had learned so much, we could talk turkey like you wouldn't believe.
He explained Null move search to me with such clarity, I wished I had had time to implement it, but I had production schedules...lots of them.
ChessChallenger
Member
Posts: 50
Joined: Mon Dec 21, 2015 7:48 pm

Chess Challenger 1

Post by ChessChallenger »

So I built the first 1,000 Chess Challengers.

Fidelity Electronics, Ltd, was in a building on Diversey Ave on the Northside of Chicago, and had the hearing aid import company and the Bio-medical (VA sole customer) business in it. And Sid gave me a sales office to use, once he hired me away from my job at Zenith Radio Corporation where I designed IF amplifier circuits for color televisions. The CES was a success, we had orders, and he came into my office, and said he had just rented the building across the street, and the Chess Challengers would be produced there, and I was to run the production line.

Jim Clovis, the technician and handy man, from Fidelity's main building bought & installed in my building, a larger production wave solder machine than he used for the bio-medical low volume productions. He trained a guy "off the street" to run it and told me good luck and went back to his building. I taught people to stuff circuit boards, then wave soldered, then assembled, tested and packed. So I opened the building and closed it at the end of the day. I built them....and taught "techs" to repair the ones that didn't work.
Fortunately they sold and Sid went for 10,000 pieces and hired Bob Heekin, an experienced electronics production manager to continue real organized production and handle the people problems.
Then Sid rented another building down on the next street for me and I started CC3, CC10 and Checker Challenger and hired 2 engineers and a programmer to help with Bridge, Checkers and Chess.
User avatar
Steve B
Site Admin
Posts: 10140
Joined: Sun Jul 29, 2007 10:02 am
Location: New York City USofA
Contact:

Post by Steve B »

ChessChallenger wrote:


Kathy, in charge of evaluation functions with Chess Masters in her office advising her, was looking for some other blue sky approach.

But it was my watching the games and the PRVs like you do, and we did at ACM tournaments that started me asking questions of Dan. Why can we not generate checks in the quiescent search? He said because it would blow up the search and slow down. Ok, I said, but what if we only generated checks that didn't occur as often, like a knight check that forked a major piece. He said, ummm,,, that would not take much and the search would not blow up.
So that is how we slowly started developing a tactical quiescent search that had all of the things a strong chess player explores when thinking of a tactical position. But I would see that the PRV was missing these obvious strong player "tricks" and have Dan look to see if he could add them.

Because of the attack map, all this type of information was easily divined. At the Micro Tournament in Spain, it was music to my ears to have the Chess Master commentator, perhaps Mike Valvo, say The Fidelity unit was playing moves it had never before been capable of playing. Just like the Masters we played to get the certified rating, who were amazed.

I used this same type of tactical threat generation on the H8 machine, since I had attack maps with the needed information.
Hi Ron
welcome back
very interesting to read that you consulted with Dan Spracklen on some of his Fidelity programs
as you can see we are now debating whether or not to list your name ..alone ..as the author of those Excalibur computers you listed .. or.. along side Kaufmans due to his consultation with you and his opening book contributions
i know you already agreed to share the spotlight but i think the community at large ..if it wants to be fair ..will need to revisit every program author listing for every chess computer and give credit for every meaningful contribution made to every program author
of course this will not happen although in your case i think the listing will probably wind up being shared on some web sites due to your modesty and high level of integrity in sharing the credit
one web sites so far is listing you separately while another is showing a shared listing

Best Regards
Steve
Last edited by Steve B on Tue Dec 29, 2015 6:23 am, edited 1 time in total.
User avatar
Steve B
Site Admin
Posts: 10140
Joined: Sun Jul 29, 2007 10:02 am
Location: New York City USofA
Contact:

Post by Steve B »

ChessChallenger wrote:So I built the first 1,000 Chess Challengers.


Fortunately they sold and Sid went for 10,000 pieces and hired Bob Heekin, an experienced electronics production manager to continue real organized production and handle the people problems.
Then Sid rented another building down on the next street for me and I started CC3, CC10 and Checker Challenger and hired 2 engineers and a programmer to help with Bridge, Checkers and Chess.
just to be clear here..
you sold 1000 Chess Challengers and then went on to produce 10,000 more products ( such as the CC3.CC10 etc..etc..)
you did not produce more then 1000 Chess Challengers(CC1's)
is that correct?

Asking While The Iron Is Hot Regards
Steve
ChessChallenger
Member
Posts: 50
Joined: Mon Dec 21, 2015 7:48 pm

Chess Challenger 1

Post by ChessChallenger »

just to be clear here..
you sold 1000 Chess Challengers and then went on to produce 10,000 more products ( such as the CC3.CC10 etc..etc..)
you did not produce more then 1000 Chess Challengers(CC1's)
is that correct?

Asking While The Iron Is Hot Regards
Steve

The Chess Challenger hardware PCB was a disaster.
The Tech at some university who layed out the PCB for Fidelity's Bio Medical division had no clue about PCB layout. God bless the guy, he hand wired the prototype Chess Challenger. Hand wired...and it worked.

I worked on electronics at Zenith in the 45Mhz IF group. I learned about good PCB layout.
So there was no way I was letting us go beyond the first 1,000 PCB.
I talked to Fidelity's talented mechanical designer to see if he would do it with my advice and explained the black tape we used at Zenith. He looked into it and said they use blue and red translucent tape for double side PCBs. So I showed how to layout ground and power first, with heavy traces and ground copper area.

He did it and it was beautiful, and reliable.
Those were the PCBs used in the next production. I am not sure when I finished, but I started working on the CC3 in my office while in the production building while the automated resistors machine was stuffing PCBs.

I don't know if the CC3 ROMs started with the new good PCB or not.
But there were only 1,000 of those horrible PCBs.
Brian B
Member
Posts: 74
Joined: Mon Jun 09, 2014 10:37 pm

Post by Brian B »

ChessChallenger wrote:So I built the first 1,000 Chess Challengers.

Then Sid rented another building down on the next street for me and I started CC3, CC10 and Checker Challenger and hired 2 engineers and a programmer to help with Bridge, Checkers and Chess.
I purchased CC3 while I was still in high school, I think it cost me around $275 or so. That was an enormous sum of money for me those days and I still remember the excitement when the computer arrived. Finally, I could play a game of chess any time I wanted to play! Or play at all, this was in 1978 and in those days I had a hard time finding a game around the neighborhood. While CC3 wasn't as strong as I hoped it would be, it still beat a few of the players on the high school team, and if I wasn't careful I would drop a piece to it now and again. As I remember it, I think Level 2 was effectively tougher to beat than Level 3 for some reason. If you hung a piece, it would definitely find it on Level 2.

These days one of my favorite computers is the Excalibur GM. It is just the right strength for me as I tend to score around 40% or so on it.

Ron, many thanks for your contributions! I hope you find the time to get back into computer chess once again.

All the best,
Brian B
User avatar
spacious_mind
Senior Member
Posts: 3999
Joined: Wed Aug 01, 2007 10:20 pm
Location: Alabama
Contact:

Post by spacious_mind »

Hi Ron,

Many thanks for returning to us and a belated Merry Christmas to you. Many of us are trying to correct our Webpages as a result of all the information that you have provided, and I am sure that we will have to rewrite a lot more as a result of all the new information that you are giving us.

In addition to correcting the computers with your Authorship, I have been trying to also correct the hardware where possible as well based on the information you provided in your computer listing. Hopefully you can recall some hardware details to help us correct any mistakes we might be making.

1997 Excalibur Chess 6805 4K Chess 7/23/1997
1997 Sabre 6805 4K Chess 7/23/1997
1997 Squire 6805 4K Chess 7/23/1997
1997 Kingmaster II 6805 Chess & Checkers 7/23/1997
1997 Travel Kingmaster II 6805 Chess & Checkers 7/23/1997
1997 Crusader 6805 Chess & Checkers 7/23/1997

In the above list I can follow all the computers except for Sabre as it seems to be a computer from pre 1995. Sabre II I think came out in 06/95 based on a time sticker on the computer (which may also be a meaningless sticker). You do have Saber III missing on your list however that should fit nicely into your timeline above, especially since its design is similar to Crusader that matches your timeline. Is it possible that Saber III Model 901-E3 is meant under Sabre?

Image

Based on your invaluable information we did find a spec sheet for the 6805 and have been able to fill in the hardware gaps. However as a couple of the 6805 in your list state 4K I am still not sure if the ROM should be 4 KB in size. It would be great if you could recall and confirm this.

Image

With Alexandra we are still unsure about its speed the spec shows 5 MHz but we don't know if perhaps you used at 10 MHz or 12 MHz quartz as well.

Lastly with Chess Station:

Image

Are we correct in assuming 6 MHz?

Very many of your designed computers have the above 3 ROM's therefore in getting it right on the above three computers, it will help us a lot in figuring out most of the other Excalibur computers.

Many thanks and despite our all our heated debates from passionate people that we all are, we do all have one thing in common which is that we all do love your posts and we are all honored that you are here sharing with us your past.

Many thanks and best regards
Nick
ChessChallenger
Member
Posts: 50
Joined: Mon Dec 21, 2015 7:48 pm

Post by ChessChallenger »

Brian B wrote:
ChessChallenger wrote:So I built the first 1,000 Chess Challengers.

Then Sid rented another building down on the next street for me and I started CC3, CC10 and Checker Challenger and hired 2 engineers and a programmer to help with Bridge, Checkers and Chess.
I purchased CC3 while I was still in high school, I think it cost me around $275 or so. That was an enormous sum of money for me those days and I still remember the excitement when the computer arrived. Finally, I could play a game of chess any time I wanted to play! Or play at all, this was in 1978 and in those days I had a hard time finding a game around the neighborhood. While CC3 wasn't as strong as I hoped it would be, it still beat a few of the players on the high school team, and if I wasn't careful I would drop a piece to it now and again. As I remember it, I think Level 2 was effectively tougher to beat than Level 3 for some reason. If you hung a piece, it would definitely find it on Level 2.

These days one of my favorite computers is the Excalibur GM. It is just the right strength for me as I tend to score around 40% or so on it.

Ron, many thanks for your contributions! I hope you find the time to get back into computer chess once again.

All the best,
Brian B
A customer (a consumer) of the CC1 flew down from Canada to see Fidelity and we talked. I told him the CC1 was a weak chess opponent, why did he like it so much. It was because he could play an opponent behind closed doors that would not embarrass him or ridicule him. I let him play one of the CC3 test units.

The CC1 was a 1 ply search machine with an exchange evaluator.
CC3 was a selective search program. Level 1 was the 1 Ply of CC1, 2 and 3 were different selective search criteria. My later machines always had a selective search level along with fixed ply no selective search levels.
User avatar
Steve B
Site Admin
Posts: 10140
Joined: Sun Jul 29, 2007 10:02 am
Location: New York City USofA
Contact:

Re: Chess Challenger 1

Post by Steve B »

Steve B wrote:

just to be clear here..
you sold 1000 Chess Challengers and then went on to produce 10,000 more products ( such as the CC3.CC10 etc..etc..)
you did not produce more then 1000 Chess Challengers(CC1's)
is that correct?

Asking While The Iron Is Hot Regards
Steve


ChessChallenger wrote:
I don't know if the CC3 ROMs started with the new good PCB or not.
But there were only 1,000 of those horrible PCBs.
Thanks Ron for your reply
1000 History making dedicated chess computers is a better description
:P

Best Regards
Steve
Post Reply