Critter bugfix update

[rvida]

download link: http://www.mediafire.com/file/vmpo1bm8f ... r_1.01.zip

- fixed book loading problems (was reported here)
- updated gtb32.dll and gtb64.dll files to version 0.4 of gaviota tablebase probing code (fixes en passant handling). These should be also compatible with Critter 0.90
- fixed a very nasty bug with storing tablebase probe results in hash table. Thanks to Jouni Uski for reporting this (more details here)

When not using tablebases, the node counts in every position should be same as in the original 1.0 release.

Linux version is not updated because it does not use tablebases yet.

Richard

[Dayffd]

Just started Critter 1.0's first tournament here. 5 minutes later...an update. Seriously, thank you for the update.

download link: http://www.mediafire.com/file/vmpo1bm8f ... r_1.01.zip

- fixed book loading problems (was reported here)
- updated gtb32.dll and gtb64.dll files to version 0.4 of gaviota tablebase probing code (fixes en passant handling). These should be also compatible with Critter 0.90
- fixed a very nasty bug with storing tablebase probe results in hash table. Thanks to Jouni Uski for reporting this (more details here)

When not using tablebases, the node counts in every position should be same as in the original 1.0 release.

Linux version is not updated because it does not use tablebases yet.

Richard

[AdminX]

Thanks!

[AdminX]

Okay I see TBhits when I use "everywhere". I still don't see them when I use "Only at Root". Which setting do you think is best? Only at Root or Everywhere?

[IWB]

Hi Richard,

download link: http://www.mediafire.com/file/vmpo1bm8f ... r_1.01.zip

- fixed book loading problems (was reported here)
- updated gtb32.dll and gtb64.dll files to version 0.4 of gaviota tablebase probing code (fixes en passant handling). These should be also compatible with Critter 0.90
- fixed a very nasty bug with storing tablebase probe results in hash table. Thanks to Jouni Uski for reporting this (more details here)

When not using tablebases, the node counts in every position should be same as in the original 1.0 release.

Linux version is not updated because it does not use tablebases yet.

Richard

First of all, thx for the release.

Any guess if this will have an influence in testing results, particualary IPON conditions which are "just" with 4 pc and in the Classic GUI (No CB)? Can the 1.0 substituded by this Version without any hassle?

Thanks
Ingo

[Norm Pollock]

Hi Richard,

The new version 1.01 does not work for older w32 computers (non-sse). Version 1.00 worked.

-Norm

[rvida]

Hi Richard,

Any guess if this will have an influence in testing results, particualary IPON conditions which are "just" with 4 pc and in the Classic GUI (No CB)? Can the 1.0 substituded by this Version without any hassle?

Thanks
Ingo

The change should be perceivable only in infinite analysis or with very long TC. Depths reached with IPON time control should not be affected at all.

Richard

[mhalstern]

Richard,

Thanks,

Will it have value for me to replace the Dll's, if I continue to use the original 1.0 release, or do the new Dll's (I am only using the 64 bit version) require the patched version?

[Mike S.]

I'm running Critter 1.01 32-bit here on a Pentium 4, Windows XP Professional 32-bit. Seems to run ok in several interfaces, incl. Gaviota tbs. access from USB stick.

A nice example of what Critter can do even under these low tech conditions:

128: # in 19 - (not #18)
[D]k7/Pp6/1Pp5/2Pp4/3Pp3/4P3/5P2/6K1 w - - 0 1

Analysis by Critter 1.01 32-bit+GTB:

1.f4 exf3 2.Kf1 f2 3.e4 dxe4 4.Kxf2 e3+ 5.Kf1 e2+ 6.Ke1
+- (2.26) Depth: 6/12 00:00:00 0kN
(...)
1.f4 exf3 2.Kf1 f2 3.e4 dxe4 4.Kxf2 e3+ 5.Ke1 e2 6.d5 cxd5 7.Kxe2 d4 8.Kd2 d3 9.c6 bxc6 10.Kxd3 Kb7 11.Kc4 c5 12.Kxc5
+- (#20) Depth: 10/32 00:00:00 3kN, tb=54
(...)
1.f4 exf3 2.Kf1 f2 3.e4 dxe4 4.Kxf2 e3+ 5.Ke1 e2 6.d5 cxd5 7.Kxe2 d4 8.Kd2 d3 9.c6 bxc6 10.Kxd3 Kb7 11.Kc4 c5 12.Kd5 c4 13.Kd6 Ka6 14.a8Q+ Kb5 15.b7 Kb4 16.b8Q+ Kc3 17.Qa3+ Kd4 18.Qh8+ Ke4 19.Qe5#
+- (#19) Depth: 19/42 00:00:00 107kN, tb=1384
(...)
1.f4 exf3 2.Kf1 f2 3.e4 dxe4 4.Kxf2 e3+ 5.Ke1 e2 6.d5 cxd5 7.Kxe2 d4 8.Kd2 d3 9.c6 bxc6 10.Kxd3 Kb7 11.Kc4 Ka8 12.Kc5 Kb7 13.a8Q+ Kxa8
+- (#19) Depth: 29/42 00:00:03 2049kN, tb=18171

(1.f3 and 1.f4 are equal and #19 is correct.)

[Carlos Ylich]

Thanks for your great work ! You is Great !