Rybka 4.1download!

[noctiferus]

http://rybkaforum.net/cgi-bin/rybkaforu ... #pid320768

[Tom Barrister]

It appears to be a bug-fix version. It was issued in a lightning-fast 10 months. Of course, there's no immediate money in fixing a product that people have already paid for, and the time used for fixing the bugs could probably be better spent on such activities as decompilation, obfuscating code, destruction of all source-code related to certain versions of a program, consultation with attorneys about impending and potential lawsuits, etc.

[Dr.Wael Deeb]

It appears to be a bug-fix version. It was issued in a lightning-fast 10 months. Of course, there's no immediate money in fixing a product that people have already paid for, and the time used for fixing the bugs could probably be better spent on such activities as decompilation, obfuscating code, destruction of all source-code related to certain versions of a program, consultation with attorneys about impending and potential lawsuits, etc.

[noctiferus]

Looks like not all known bugs have been fixed:
http://rybkaforum.net/cgi-bin/rybkaforu ... pid=320768
http://www.open-chess.org/viewtopic.php ... 6&start=10

[Jouni]

Even if IPON don't show any improvement, there is! In analysis/test suites Rybka 4.1 is better and solves more positions in all/most testsuites.
Also some mates are found clearly faster, may be not with correct length, though. So may be R4.1 needs longer time control to show it's teeth?

Jouni

[El Gringo]

It appears to be a bug-fix version. It was issued in a lightning-fast 10 months. Of course, there's no immediate money in fixing a product that people have already paid for, and the time used for fixing the bugs could probably be better spent on such activities as decompilation, obfuscating code, destruction of all source-code related to certain versions of a program, consultation with attorneys about impending and potential lawsuits, etc.

[mhalstern]

I don't see posts in the 2 threads that you mentioned, claiming unfixed bugs.

[noctiferus]

link1. - 1st claim. post by VolMike Date 2011-03-07 15:24
Bishop underpromotion feature is still not added...
- 2nd claim post by Pascal Pflaum Date 2011-03-08 03:59 Rybka 4.1 doesn't support "go nodes x" anymore


link 2, page 2, Jeremy Bernstein post: Subject: Re: Rybka 4.1 Update released Posted: Mon Mar 07, 2011 8:16 pm. :
png image by Dragon Mist Date 2011-03-07 13:56 from
http://rybkaforum.net/cgi-bin/rybkaforu ... ?tid=21107
and following comments
enrico

PS: sorry for being unable to include here the posts images. Anyway, I hope you will easily find the original posts.

[M ANSARI]

The biggest change in R4.1 is fixing the MPV bug, and fixing the stall bug that could occur at depth 13 and higher. As an analysis engine it is a huge improvement as without the stall bug, the evaluation of positions can finally show its true accuracy. I still would consider it a "slower" engine than the very top engines, but I believe it has the more accurate evaluation of static positions.

[michiguel]

The biggest change in R4.1 is fixing the MPV bug, and fixing the stall bug that could occur at depth 13 and higher. As an analysis engine it is a huge improvement as without the stall bug, the evaluation of positions can finally show its true accuracy. I still would consider it a "slower" engine than the very top engines, but I believe it has the more accurate evaluation of static positions.

What is specifically the stall bug?

Miguel