Simon

[Olivier Deville]

Could somebody please send me Simon ? I am unable to connect to Dann's ftp right now.

Thanks in advance.

Olivier

[Jim Ablett]

On the way ....

Jim.

[Olivier Deville]

Got it, thanks

Olivier

[Dan Honeycutt]

Hi Oliver:

First: If you use Simon make sure it's version 1.1. Dann Corbit found a bug where the mate score was sign reversed which caused Simon to avoid checkmate if it found one.

Second: You probably should not use it at all. Simon is designed to be used by a beginning chess programmer as a starting point. It has all the mechanics necessary to play chess; move generator, xboard interface, fen parser etc but no brains. The search and evaluation are very basic and the program is extremely weak. This is by design - so a new programmer can easily make a few improvements and be rewarded with a significant improvement in its play. Simon is not intended for tournament play.

Best
Dan H.

[Guenther Simon]

Hi Oliver:

First: If you use Simon make sure it's version 1.1. Dann Corbit found a bug where the mate score was sign reversed which caused Simon to avoid checkmate if it found one.

Second: You probably should not use it at all. Simon is designed to be used by a beginning chess programmer as a starting point. It has all the mechanics necessary to play chess; move generator, xboard interface, fen parser etc but no brains. The search and evaluation are very basic and the program is extremely weak. This is by design - so a new programmer can easily make a few improvements and be rewarded with a significant improvement in its play. Simon is not intended for tournament play.

Best

Too late Dan How could I have resisted to include it as soon as possible
into one of my test tourneys with THAT name.
I can verify that bug now It sometimes tried to miss mates in one.
Seriously, I did not know it was not intended for any tournaments,
as there was no information when Dan announced it.

Best wishes,
G. Simon

[Dan Honeycutt]

Too late Dan How could I have resisted to include it as soon as possible
into one of my test tourneys with THAT name.
I can verify that bug now It sometimes tried to miss mates in one.
Seriously, I did not know it was not intended for any tournaments,
as there was no information when Dan announced it.

Hi Guenther:

I'm sorry but it wasn't named after you but rather the nursery rhyme "Simple Simon met a pie man going to the fair..."

You can of course include it in any tests or tournaments you wish, just don't expect much. Do use the 1.1 version though. While Simon is weak by design it is not intended to be so weak that it can't finish off a mate in one.

Best
Dan H.

[Olivier Deville]

Hi Dan

No way, if you release a thing, I'll take it in my next tourney !

And believe it or not, Simon will avoid the Promotion Class (formerly known as Trash Division ) and will start in my F tourney starting in April.

Olivier

[GeoffW]

Hi Dan

Thanks for releasing the source code for Simon. I had thought it would be very useful for beginners to have a TSCP equivalent but using bitboards. I hadnt come across one that looked readable prior to Simon, I might even get to understand bitboards myself now !

If you release a new version in the future, here is a little correction for you

in void PVDisplay(int score, int mark)
I think it should read
int et = (Now() - tc_start)/10;

In Arena Simon was pretending to be 10 times slower in nodes per second than it really is. Not going to fool me that easily

regards Geoff

[Anonymous]

Hi Oliver:

First: If you use Simon make sure it's version 1.1. Dann Corbit found a bug where the mate score was sign reversed which caused Simon to avoid checkmate if it found one.

Best
Dan H.

Hi Dan

I am unable to reach Danns ftp site. Is the mate score bugfix the only change between version 1.0 and version 1.1?

Regards
Dave

[Guenther Simon]

Hi Dan

I am unable to reach Danns ftp site. Is the mate score bugfix the only change between version 1.0 and version 1.1?

Regards
Dave

But it is an essential one, as it cripples Simon completely.
Dan did not mention that the reverse mate score not only
led to avoid mates for itself, but also to _allow_ easy mates
against it! (e.g. mate in 1 or 2 already in move 14...searching
for self mates you can say

Regards,
Guenther

[Anonymous]

Hi Dan

I am unable to reach Danns ftp site. Is the mate score bugfix the only change between version 1.0 and version 1.1?

Regards
Dave

But it is an essential one, as it cripples Simon completely.
Dan did not mention that the reverse mate score not only
led to avoid mates for itself, but also to _allow_ easy mates
against it! (e.g. mate in 1 or 2 already in move 14...searching
for self mates you can say

Regards,
Guenther

I had already version 1.0 of Simon from Dan for a few weeks, and i noticed the strange mate behaviour immediately. The reverse mate score bug was found and fixed. I added check extensions, and Simon finds mates very well now.

Regards
Dave

[Guenther Simon]

I had already version 1.0 of Simon from Dan for a few weeks, and i noticed the strange mate behaviour immediately. The reverse mate score bug was found and fixed. I added check extensions, and Simon finds mates very well now.

Regards
Dave

I will test the new fixed version 1.1 soon, because I hope it is not
_too strong_. If it is stronger than expected I fear lots of 'Simons'
to come with only slight modifications from programming newbies.

IMHO it is no good idea to offer a source completely for free to anyone,
at least if it is _not_ that weak as Dan currently suggests.

I know that good programmers qualify weak programs as ones
below 2200 perhaps, but there are surely >100 selfmade programs
around, which are under 2200 and I respect their original work
a lot!

If Simon already has 2000 all silly and simple TSCP cloners will just
clone Simon now and computer chess tourneys at lower level
will be a waste of time, my two cents.

Regards,
Guenther

[Leo Dijksman]

I had already version 1.0 of Simon from Dan for a few weeks, and i noticed the strange mate behaviour immediately. The reverse mate score bug was found and fixed. I added check extensions, and Simon finds mates very well now.

Regards
Dave

I will test the new fixed version 1.1 soon, because I hope it is not
_too strong_. If it is stronger than expected I fear lots of 'Simons'
to come with only slight modifications from programming newbies.

IMHO it is no good idea to offer a source completely for free to anyone,
at least if it is _not_ that weak as Dan currently suggests.

I know that good programmers qualify weak programs as ones
below 2200 perhaps, but there are surely >100 selfmade programs
around, which are under 2200 and I respect their original work
a lot!

If Simon already has 2000 all silly and simple TSCP cloners will just
clone Simon now and computer chess tourneys at lower level
will be a waste of time, my two cents.

Regards,
Guenther

Hi Guenther,

I completely agree with your opinion above!
Looking the the number of clones of copyrighted sources found in the past and probably still excist(?) it is easy to get an idee how many renamed "Simon" engines we can expect!?

Well, I am looking to an way to avoid to have a lot of copied Simon engines playing here, the easiest way seems to add an bottom limit of 2300 elo but I dont like the idee to reject engines which are written from scratch!!

Leo.

[Pradu]

I completely agree too. I'm all up to people giving ideas out on techniques for , but I think its wrong that people steal code and be unfair to the people who spent many days programming and debugging whatever they are making.

The source code that people release is only for other people to see how something is done if the have no idea how to do it before.

And please don't reject my pathetic engine . There must be a better way to get rid of clones.

[Dan Honeycutt]

Hi All:

Ill try to answer the several questions, concerns posted above.

Olivier: I thought "trash" was rather harsh, but whatever you call it Simon will probably work its way there. I've also changed my mind - it probably is good to include it in a few tournaments to see where it rates. See my comments below.

Geoff: Thanks for the nps fix. Simon is not intended to be a bitboard TSCP. First, Simon is public domain to avoid problems that you know well. Second, TSCP as well as Faile, Gerbil and others are designed to play pretty well within the constraint that they are simple programs. Simon is not designed to play well - it is a bitboard infrastructure with only enough search and eval to show how those work.

Dave: The mate score bug is the only real difference. Dann was going to add some clarification/readability fixes such as parens so some expressions are clear without relying on order of precedence etc. He may incorporate Geoff's nps fix which is fine. But I specifically asked him not to add anything directed at playing strength. BTW - I assume you solved the problem you had with your compiler (?)

Several: How strong is Simon? Some tournaments will tell. It has nothing to intentionally cripple it like ply, nodes or search time limitations. But it also has nothing to intentionally make it play better like null move, extensions, pruning, IID etc. But it's not weak in the sense that it would be a good opponent for someone new to chess. Any program that is fairly bug free on modern hardware will play at a 2000+ level.

Guenther, Leo and other TD's:

I am concerned about a flock of weak Simons diluting the quality of tournaments. Note in the readme that the only request I make to anyone using Simon is that they achieve significant improvement before they enter it in any tournament. But I think that is why you have multiple divisions and there be an open class, trash class or whatever you call it. I don't think excluding engines below xxxx is a good approach because you exclude new programmers (one of whom may have SMK's gifts) who just haven't gotten very far yet.

The clone situation is a nasty problem that has no complete solution. My hope with Simon is to help, not make the problem worse. I think there are 2 types of people who clone or use others source. First, those who only want to win and don't mind taking someone else's work and calling it their own. Second, those who have a genuine interest in chess programming, don't want to steal anything, but find the task of just getting started overwhelming. Simon is intended for the second type, the first type are going to use Crafty, Fruit, Pepito, Phalanx or other strong open source program.

I don't want to take the thread into a closed-source vs open-source debate - there are pros and cons for both sides. I have always been in the open-source camp as a means to share ideas and help others (as others have helped me). The source for future releases of Bruja will only be available to other established engine authors or computer chess contributors following Tord's swapware idea. Bruja is not as strong as the above named programs, but it is strong enough that it could be a candidate for someone (of the first type) to clone. I hope to keep the source out of their hands. To not leave the beginner out in the cold I created Simon.

The distinction I made above between Simon and TSCP and other simple programs is, I think, important. Simon is not intended to "play pretty well for a simple program". It is merely the infrastructure, the "nuts and bolts" needed for a chess program to play. For a program to play well requires an efficient search and a decent evaluation. Those functions are included in Simon in almost skeletal form - just enough to show they work. Whatever level of play any Simon derivative achieves is the work of the author, not me.

Best
Dan H.

[Anonymous]

Hi All:
Dave: The mate score bug is the only real difference. Dann was going to add some clarification/readability fixes such as parens so some expressions are clear without relying on order of precedence etc. He may incorporate Geoff's nps fix which is fine. But I specifically asked him not to add anything directed at playing strength. BTW - I assume you solved the problem you had with your compiler (?)
Best
Dan H.

Thanks Dan. Do you know if Dann added the clarification/readability fixes in version 1.1 or if they will be in a future version? I still can't reach his ftp site, so i'm unable to download version 1.1.

And yes, i did solve my compiler problem, thanks for asking.

Regards
Dave

[Dan Honeycutt]

Thanks Dan. Do you know if Dann added the clarification/readability fixes in version 1.1 or if they will be in a future version? I still can't reach his ftp site, so i'm unable to download version 1.1

Hi Dave:

Look for 1.2 which will have any cleanup Dann does plus the nps fix that Geoff mentioned above. It won't play any different than 1.1 and it will be a few days before it's available. Dann or Dan will let you know.

Best
Dan H.

[Dann Corbit]

Is avaiable now on my FTP site.

By now, you know where to look, I imagine.
But just in case,
ftp://cap.connx.com/pub/chess-engines/n ... /simon.zip

This one uses the book of Bruja. It is easy to turn it off.
One with bigger teeth might be forthcoming.