Fast Improvement of Colossus

[fern]

Last Colossus version is already in the top ten league in the test done by one pal here. Considering the fact that Bryant begun to put his hands on Colossus again just one year ago or something, after many years when he did nothing with this old program, 80's vintage, it is very remarkable and a show that age is awful for us, average people, not for talented people.
I am sure Colossus will be soon in the top five league AND that Bryan even will offer us a new, gorgeous gui.
Just wait.

Ferdinan

[Tony Thomas]

I am surprised to see that an engine at Colossus' level is making that kind of improvements. I wish it was possible to do the same for my all time favorite engine, crafty.

[Dann Corbit]

The reason for the fast advance of Colossus is the same as the reason for the fast advance of Movei and Alaric.

Smart dedicated people put effort into their programs and the results come out well.

The biggest problem with chess programming is that sometimes the programmers get bored with it. That happened to Zappa for a while and some other up and coming engines like Thinker.

[Uri Blass]

The reason for the fast advance of Colossus is the same as the reason for the fast advance of Movei and Alaric.

Smart dedicated people put effort into their programs and the results come out well.

The biggest problem with chess programming is that sometimes the programmers get bored with it. That happened to Zappa for a while and some other up and coming engines like Thinker.

I am unsure if it happened to zappa.

Even when Anthony claimed that he was retired zappa got a significant improvement.

I do not know what he did but even if he did no almost coding after zap!chess paderborn and only made automatic testing to find a better personality then I cannot consider it as being retired.

Uri

[Martin Thoresen]

I am unsure if it happened to zappa.

Even when Anthony claimed that he was retired zappa got a significant improvement.

I do not know what he did but even if he did no almost coding after zap!chess paderborn and only made automatic testing to find a better personality then I cannot consider it as being retired.

Uri

Uri, I am curious, why are you so obsessed with this whole thing about Anthony being retired or not?

Why don't you cut Anthony some slack and acknowledge his achievements instead.

With that said, I also said I'll retire from computer chess, but alas, something pulled me back in.

Now just ease up and appreciate the current computer chess situation instead.
I'm sure I'm not the only one smiling at the current Zappa - Rybka match...

[Wardy]

The endgame test suite result posted elsewhere on the site is also very impressive beating some other big names, though still with some way to go to challenge the very best.

I just hope that Martin can continue the progress he's made so far.

[fern]

Probably you are right. It is sad that Withington is one of those that got bored, as much his Cstal was a real promise. Martin, instead, caught again the virus and working without anxieties of any kind, just for fun, he is makings great strides to recover lost time. He was in the 80's a remarkable programmer and much of that and new stamina and knowledge helps him now.

My bet:
Colossus will be in the 5 top league in around a semester.
My best.
Fernando

[Graham Banks]

Colossus 2007d hasn't started too well for me in my testing, but it's still early days of course!

[GS]

Last Colossus version is already in the top ten league in the test done by one pal here. Considering the fact that Bryant begun to put his hands on Colossus again just one year ago or something, after many years when he did nothing with this old program, 80's vintage, it is very remarkable and a show that age is awful for us, average people, not for talented people.
I am sure Colossus will be soon in the top five league AND that Bryan even will offer us a new, gorgeous gui.
Just wait.

Ferdinan

Colossus is a fine program with steady progress on high level and my
thanks goes to Martin, but...
...the little test you have seen was just a tiny endgame test, which also
lacks a lot of programs from the top. It is no top 10 program yet.

Guenther

[Wardy]

I hope it will pull through

I tested a few versions in the run up to 2007d and each on it's own was slightly better the 2007c so I'd hope that all the changes together should show an improvement.

Though the proof of the pudding...............