Rookie still under active development

[Tony Thomas]

I havent heard of some of these engines that competed in Australian computer chess championship. May be Ross Boyd can tell you if the ones we have never heard of actually supports winboard protocol.

Fencer
Shaun Press
Fencer is a follow up program to Vanilla Chess. It uses bit fields to generate moves and evaluate mobility. It has pretty fast evaluation but a slower move generator. A big kingsafety function ecourages it to launch hack attacks which work well against humans but not against other computers.

JA hooked up Vanilla chess by Shawn Press, it was at the level of TSCP at the most.

Decapod

Tor Lattimore
Alpha-Beta search, PVS, hash tables, null-moving, killers, history, opening book, 0x88 move-gen with pawn attack tables. ICC, 2400 blitz, 2100 standard. Pretty weak program. Blitz doesn't mean much on ICC, 2400 is
achieved by not losing pieces too quickly. Maybe about 1600 ACF? Has a 3D opengl board (but that uses a fair bit of CPU)
P IV, 2.8 gHz, 512 MG ram.

Decapod is probbably an earlier version of K9??

Desperado
David Blackman
Desperado has been under development since 1989. It has played in most of the
NC3 tournaments, and all the ANU Carbon-Silicon tournaments before that.
Highlights include a win against IM Guy West the year he became Australian
Champion, and 2 third places in NC3.

Code size is about 7500 lines, which is probably bigger than a chess program
should be. Desperado has a few extensions, heavy null-move pruning, hash
tables, better speed than most NC3 entrants on similar hardware, a moderately
complex evaluation function, and no opening book.
Pentium-2 Celeron at 533MHz, running Linux

[Ron Murawski]

I havent heard of some of these engines that competed in Australian computer chess championship. May be Ross Boyd can tell you if the ones we have never heard of actually supports winboard protocol.

Fencer
Shaun Press
Fencer is a follow up program to Vanilla Chess. It uses bit fields to generate moves and evaluate mobility. It has pretty fast evaluation but a slower move generator. A big kingsafety function ecourages it to launch hack attacks which work well against humans but not against other computers.

JA hooked up Vanilla chess by Shawn Press, it was at the level of TSCP at the most.

Decapod

Tor Lattimore
Alpha-Beta search, PVS, hash tables, null-moving, killers, history, opening book, 0x88 move-gen with pawn attack tables. ICC, 2400 blitz, 2100 standard. Pretty weak program. Blitz doesn't mean much on ICC, 2400 is
achieved by not losing pieces too quickly. Maybe about 1600 ACF? Has a 3D opengl board (but that uses a fair bit of CPU)
P IV, 2.8 gHz, 512 MG ram.

Decapod is probbably an earlier version of K9??

Desperado
David Blackman
Desperado has been under development since 1989. It has played in most of the
NC3 tournaments, and all the ANU Carbon-Silicon tournaments before that.
Highlights include a win against IM Guy West the year he became Australian
Champion, and 2 third places in NC3.

Code size is about 7500 lines, which is probably bigger than a chess program
should be. Desperado has a few extensions, heavy null-move pruning, hash
tables, better speed than most NC3 entrants on similar hardware, a moderately
complex evaluation function, and no opening book.
Pentium-2 Celeron at 533MHz, running Linux

Hi Tony,

If Tor Lattimer is competing in the 2006 Australian chess championships using Decapod, then I think that would be his latest engine. his K9 engine was playing at Ridderkerk in 2005. I'll list his program as Decapod [K9?].

Thanks for your additions and corrections!

Ron