Hello all

[S.Taylor]

I would say about 15 years. I worked for Mentor Graphics for 12 1/2 years and have been retired 2 1/2 years now.

Probably after making the Novag Diamond, a second one, i think.

[mephisto]

Novag told me Dave's last update was for the Star Diamond in 2003

[mephisto]

Dave
If you do decide to add your own code to Crafty, will it then be called WCrafty

[Kingghidorah]

Nice to see Dave back, the good ole' wChessx days,ahhh. Thanks for putting up WChess. I remember using it a lot vs crafty & ferret.

[DaveKitt]

Had to research PSH as I suffer from TMA (too many acronyms). 'pre search heuristics' involved recognizing certain specific characteristics and feeding bonus/penalties into specific moves in the root position move list (ply one more list). Most famous example of course is the bxh7 sac, although as I recall there were a couple for center pawn sacs when danger of getting constricted to death. These rules were done from white POV then list run through and swapped for black POV. With limited processing and search depth, at the time it seemed necessary to look for some typical chess themes cheaply (applied only for the plyone moves) as opposed to testing lots of positional features deeper in the search. Obviously there are all kinds of negatives to this approach, but in practice worked out reasonably enough.

As I recall, another short cut which led to interesting play at times had to do with long strings of checks which were not repeats. With Super Conny at least, these were given a bonus as this was before all the more intelligent extensions for checks had been worked out (and I think ram space only allowed something like 18 ply maxdepth). Idea was that such strings of checks might be leading to forced mate or gain of material which program did not have depth to see actual end of.

[DaveKitt]

I think I only provided them w/one of the old programs which they had misplaced, don't recall I did any development work at that late date. They had a number of programs which ran on different CPUs and the i/o for most was rather straight forward to move to other, electronically similar platforms, even though form factors might be quite different.

[DaveKitt]

I need to open the bidding...should I use Stockfish or Crafty as the base?
Seriously, I don't think I would ever write another chess program, although I was taking a nice walk down memory lane reading the version log in Crafty's main.c. So many of same problems seen and solved back in the day. So many, many times trying something, modifying, rejecting, going back, tweaking....then finding small bug that made lots of the test results questionable and going back through the whole process....

[JuLieN]

I need to open the bidding...should I use Stockfish or Crafty as the base?
Seriously, I don't think I would ever write another chess program, although I was taking a nice walk down memory lane reading the version log in Crafty's main.c. So many of same problems seen and solved back in the day. So many, many times trying something, modifying, rejecting, going back, tweaking....then finding small bug that made lots of the test results questionable and going back through the whole process....

Some went the Ippolit way instead, because Ippolit is totally free of license
http://ippolit.wikispaces.com/

(I can't believe I'm suggesting some new cloning operation! )

But IMO, if you were to start from other engines' sources you'd probably lose some prestige... Except if you manage to get it top the charts.

[mclane]

i always had the feeling that PSH worked very well for Super-Constellations,
but they were too many of them in the Expert.

i thought: why isn't he continuing this in the later machines with better hardware, such as the 6301 machines (supremo 8mhz) or the later h8 machines with up to 26 mhz.

also when novag was in the crisis later, and peter auges daughter
created the 2nd ROBOT version, i wondered: why don't they put the Super-Conny program into it. instead they did a weak and strange program into it, that played lousy.

the super-constellation days were very good. we all miss those sacrifices of the super-constellation. or when 3 or 4 pieces on board were hanging, and you could not eat ANY of them because each one would have been mate or loss. THAT was nice.

i always looked for the super-constellation roots in later novag machines.
thought maybe he put this approach back ...

here some examples of the famous super-constellation games:

http://www.schachcomputer.info/forum/f1 ... -2599.html

or the famous game:

[pgn]
[Event "?"]
[Site "?"]
[Date "1984.06.01"]
[Round "?"]
[White "Super-Constellation"]
[Black "Elite A/S 4.6 mhz"]
[Result "1-0"]

1. e4 c5 2. Nf3 Nc6 3. d4 cxd4 4. Nxd4 g6 5. Nc3 Bg7 6. Be3
Nf6 7. Bc4 Qa5 8. O-O O-O 9. Bb3 d6 10. h3 Bd7 11. f4 Nxd4
12. Bxd4 Bc6 13. Nd5 Rfe8 14. f5 Nxe4 15. fxg6 hxg6
16. Bxg7 Nd2 17. Ne3 Nxf1 18. Qxf1 Kxg7 19. Qxf7+ Kh6
20. Ng4+ Kg5 21. Qg7 Kh4 22. Qxg6 Qb4 23. g3+ Kxg3 24. Nh6+
Kxh3 25. Be6+ Kh4 26. Nf5+ Kh3 27. Qg3#
[/pgn]

[JuLieN]

i always had the feeling that PSH worked very well for Super-Constellations,
but they were too many of them in the Expert.

i thought: why isn't he continuing this in the later machines with better hardware, such as the 6301 machines (supremo 8mhz) or the later h8 machines with up to 26 mhz.

also when novag was in the crisis later, and peter auges daughter
created the 2nd ROBOT version, i wondered: why don't they put the Super-Conny program into it. instead they did a weak and strange program into it, that played lousy.

the super-constellation days were very good. we all miss those sacrifices of the super-constellation. or when 3 or 4 pieces on board were hanging, and you could not eat ANY of them because each one would have been mate or loss. THAT was nice.

i always looked for the super-constellation roots in later novag machines.
thought maybe he put this approach back ...

@Dave

Thorsten's message makes me wonder: if PSH works well with shallow searches, why not apply it in deeper tree, at, say, horizon-6 ? Each node of the tree is the root of its sub-tree, so if PSH was efficient with the 5-6 plies deep trees these old hardwares could generate, then maybe it would still work in today's programs as well, but later in the tree?