Winboard Forum

Posted: **24 Feb 2005, 11:32**

Anthony send me a new version of his program. Also the book has been updated. This version works on my Athlon and I'm pretty sure it will work on others too. However, please make sure the program works for you before you download the 13 MB book.

Release notes:

1. Major bug in the book editor fixed; thresholding now works correctly.

2. Exactly the version that competed in CCT.

3. SHOULD work on Athlon CPUs. Should is the operative word here.

4. Definitely stronger than Zappa 0.8B2. I am not sure how much; my guess is 50 elo. But you guys would get bored without mysteries in life, wouldn't you? I am personally more interested in slow timecontrols, but feel free to test Zappa at any speed.

5. Consider this line an official request that Zappa be entered in any tournament available.

6. If config files show up in notepad without line breaks, open them with wordpad.

7. Have fun!

http://volker-pittlik.name/zappa/zappa.html

Regards

Volker

P.S. I'm not sure what should happen with the 64-bit version.

Posted: **24 Feb 2005, 12:36**

It exits on Celeron 1GHz

From log file:
set_option(): threads = 1
game_start_thread(0): spawning worker thread!
error starting workers: 87

Igor

Posted: **24 Feb 2005, 12:43**

It works on this P3 on which it did not work before!

Posted: **24 Feb 2005, 12:57**

Igor Gorelikov wrote:It exits on Celeron 1GHz

From log file:
set_option(): threads = 1
game_start_thread(0): spawning worker thread!
error starting workers: 87

Exactly the same behaviour here (Win98, Celeron/400)!

I give it up ...

Franz.

Posted: **24 Feb 2005, 13:15**

F.Huber wrote:
Igor Gorelikov wrote:It exits on Celeron 1GHz

From log file:
set_option(): threads = 1
game_start_thread(0): spawning worker thread!
error starting workers: 87

Exactly the same behaviour here (Win98, Celeron/400)!

I give it up ...

Franz.

As far as I know Celeron A processors don't have SSE instructions which Zappa seems to use, so I'm afraid the only hope for you is that Anthony will make a compile for older processors. :-(

Posted: **24 Feb 2005, 14:49**

Zappa exits on my Athlon 1800XP (Win98, 256 Mb, 1,53Ghz)

logfile says:

set_option(): tbpath = c:/tb
tablebase_init(): searching for EGTBs at path "c:/tb"
tablebase_init(): found EGTBs with up to 5 pieces.
tablebase_init(): EGTB indices & decompression using 3558kb of memory
set_option(): phash = 8388608 (8M)
hash_reallocate(): Allocated transposition table: 64M
set_option(): threads = 1
game_start_thread(0): spawning worker thread!
error starting workers: 87

Best greetings

Hartmut Woldeit

Posted: **24 Feb 2005, 14:58**

Fantastic! - Well done Anthony.

Zappa now runs on my Athlon XP-M /WinXp.
Here's my little contribution to celebrate this
and your recent tournament triumph.

Thanks,

Jim.

Posted: **24 Feb 2005, 15:08**

I think by the above posts that the problem is related to Win98. Zappa runs fine under Windows XP.

Igor

Posted: **24 Feb 2005, 15:20**

F.Huber wrote:
Igor Gorelikov wrote:It exits on Celeron 1GHz

From log file:
set_option(): threads = 1
game_start_thread(0): spawning worker thread!
error starting workers: 87

Exactly the same behaviour here (Win98, Celeron/400)!

I give it up ...

Franz.

The same behaviour here (WinMe, P3) :-(

Dave

Posted: **24 Feb 2005, 15:22**

Volker Pittlik wrote: However, please make sure the program works for you before you download the 13 MB book.

Speaking of the large book... any news on whether it has been fixed too?

Cheers,

Martin

Posted: **24 Feb 2005, 15:48**

Martin Baumung wrote:
Volker Pittlik wrote: However, please make sure the program works for you before you download the 13 MB book.

Speaking of the large book... any news on whether it has been fixed too?

Cheers,

Martin

Works OK for me.

Code: Select all: Probed book successfully (20 entries / 1196 bytes / 114 ms). Selected move e2-e4 Nb1-c3 0.0% ?? [ 54.3%, +520, -410, =354] e2-e3 0.0% ?? [ 37.6%, +27, -51, =19] b2-b4 0.0% ?? [ 41.1%, +64, -109, =81] g2-g3 0.0% [ 55.0%, +80, -58, =83] c2-c3 0.0% ?? [ 42.7%, +11, -17, =13] a2-a3 0.0% ?? [ 52.1%, +24, -21, =25] d2-d3 0.0% ?? [ 40.6%, +17, -26, =5] g2-g4 0.0% ?? [ 32.6%, +5, -13, =5] h2-h3 0.0% ?? [ 16.7%, +0, -2, =1] f2-f3 0.0% ?? [ 0.0%, +0, -1, =0] h2-h4 0.0% ?? [ 0.0%, +0, -2, =0] b2-b3 0.7% !! [ 50.9%, +221, -210, =208] f2-f4 1.2% !! [ 44.0%, +444, -622, =416] e2-e4 98.1% !! [ 54.9%, +29174, -21806, =24930] Ng1-f3 0.0% ?? [ 55.3%, +1080, -770, =1075] c2-c4 0.0% [ 55.4%, +1045, -738, =1051] d2-d4 0.0% [ 55.0%, +6070, -4381, =6277] Ng1-h3 0.0% ?? [100.0%, +1, -0, =0] a2-a4 0.0% ?? [ 0.0%, +0, -0, =0] Nb1-a3 0.0% ?? [ 0.0%, +0, -0, =0]

One has reported a problem after download. I can't reproduce that.

It seems a bit strange that Zappa (using the big book) plays every time 1.e4 e6 2.d4 d5 3.exd5 (French Exchange) as White. This moves seems to lead to very drawish games.

Volker

Posted: **24 Feb 2005, 15:57**

David Dahlem wrote:
F.Huber wrote:
Igor Gorelikov wrote:It exits on Celeron 1GHz

From log file:
set_option(): threads = 1
game_start_thread(0): spawning worker thread!
error starting workers: 87

Exactly the same behaviour here (Win98, Celeron/400)!

I give it up ...

Franz.

The same behaviour here (WinMe, P3)

Dave

Well, it seems that we Win98/ME-users are a suppressed minority! :evil:

Nevertheless, if I?ve to decide, whether to throw away my (since more than 5 years working) Win98 or this crazy Zappa, I?m sure you can guess how my decision will be ... :wink:

Regards,
Franz.

Posted: **24 Feb 2005, 16:10**

OK, CreateThread (which I am using) has changed since Win98. One of the parameters can be NULL in Win2000 but not Win98/ME/etc. This should be simple to fix. I'll try to get it done tonight.

Really, though, no one should use those operating systems. Win2000 is just SO much better . . .

Volker: Please don't judge my booker by the public book

Arturo saved all the good lines for the tournament book, at my request.

anthony

Posted: **24 Feb 2005, 16:11**

Oh, and thanks for the logo, Jim

anthony

Posted: **24 Feb 2005, 16:33**

It seems a bit strange that Zappa (using the big book) plays every time 1.e4 e6 2.d4 d5 3.exd5 (French Exchange) as White. This moves seems to lead to very drawish games.

Volker

Volker: Please don't judge my booker by the public book Smile Arturo saved all the good lines for the tournament book, at my request.

anthony

Hi Anthony, it was me who noticed this, because I played a few Blitz
as Human vs. Zappa and may be I will even do a live Blitz match
against it, as long as Zappa always plays the French exchange vs. me.
I feel I have good chances then to win a 10 games match, when having
always Black

Will Zappa play with this book in RWBC?

Best wishes,
Guenther

Posted: **24 Feb 2005, 16:46**

Volker Pittlik wrote:
Works OK for me.
Code: Select all
Probed book successfully (20 entries / 1196 bytes / 114 ms). Selected move e2-e4 Nb1-c3 0.0% ?? [ 54.3%, +520, -410, =354] e2-e3 0.0% ?? [ 37.6%, +27, -51, =19] b2-b4 0.0% ?? [ 41.1%, +64, -109, =81] g2-g3 0.0% [ 55.0%, +80, -58, =83] c2-c3 0.0% ?? [ 42.7%, +11, -17, =13] a2-a3 0.0% ?? [ 52.1%, +24, -21, =25] d2-d3 0.0% ?? [ 40.6%, +17, -26, =5] g2-g4 0.0% ?? [ 32.6%, +5, -13, =5] h2-h3 0.0% ?? [ 16.7%, +0, -2, =1] f2-f3 0.0% ?? [ 0.0%, +0, -1, =0] h2-h4 0.0% ?? [ 0.0%, +0, -2, =0] b2-b3 0.7% !! [ 50.9%, +221, -210, =208] f2-f4 1.2% !! [ 44.0%, +444, -622, =416] e2-e4 98.1% !! [ 54.9%, +29174, -21806, =24930] Ng1-f3 0.0% ?? [ 55.3%, +1080, -770, =1075] c2-c4 0.0% [ 55.4%, +1045, -738, =1051] d2-d4 0.0% [ 55.0%, +6070, -4381, =6277] Ng1-h3 0.0% ?? [100.0%, +1, -0, =0] a2-a4 0.0% ?? [ 0.0%, +0, -0, =0] Nb1-a3 0.0% ?? [ 0.0%, +0, -0, =0]

That's funny :shock:

Here's what it looks like for me:

Code: Select all: Nb1-c3 0.0% ?? [ 54.3%, +520, -410, =354] e2-e3 0.0% ?? [ 37.6%, +27, -51, =19] b2-b4 0.0% ?? [ 41.1%, +64, -109, =81] g2-g3 0.0% [ 55.0%, +80, -58, =83] c2-c3 0.0% ?? [ 42.7%, +11, -17, =13] a2-a3 0.0% ?? [ 52.1%, +24, -21, =25] d2-d3 0.0% ?? [ 40.6%, +17, -26, =5] g2-g4 0.0% ?? [ 32.6%, +5, -13, =5] h2-h3 0.0% ?? [ 16.7%, +0, -2, =1] f2-f3 0.0% ?? [ 0.0%, +0, -1, =0] h2-h4 0.0% ?? [ 0.0%, +0, -2, =0] b2-b3 36.6% !! [ 50.9%, +221, -210, =208] f2-f4 63.4% !! [ 44.0%, +444, -622, =416] e2-e4 0.0% [ 54.9%, +29174, -21806, =24930] Ng1-f3 0.0% ?? [ 55.3%, +1080, -770, =1075] c2-c4 0.0% [ 55.4%, +1045, -738, =1051] d2-d4 0.0% [ 55.0%, +6070, -4381, =6277] Ng1-h3 0.0% ?? [100.0%, +1, -0, =0] a2-a4 0.0% ?? [ 0.0%, +0, -0, =0] Nb1-a3 0.0% ?? [ 0.0%, +0, -0, =0]

So obviously it's not related to the book-file itself

Cheers,

Martin

Posted: **24 Feb 2005, 16:54**

Martin Baumung wrote:...

That's funny

Here's what it looks like for me:

...

So obviously it's not related to the book-file itself

Cheers,

Martin

It seems you still have the old book. The new one has the following properties:

date: 2005 Feb 21, 19:52:48
size: 35'362'632 Bytes

The old one:

date: 2005 Feb 20, 10:12:28
size: 35'550'432 bytes

Volker

Posted: **24 Feb 2005, 17:29**

Volker Pittlik wrote:
It seems you still have the old book. The new one has the following properties

Oh yes, you are right Volker. Thanks!

I should have read your initial post more attentively

Martin

Posted: **24 Feb 2005, 19:18**

Guenther Simon wrote:
It seems a bit strange that Zappa (using the big book) plays every time 1.e4 e6 2.d4 d5 3.exd5 (French Exchange) as White. This moves seems to lead to very drawish games.

Volker

Volker: Please don't judge my booker by the public book Smile Arturo saved all the good lines for the tournament book, at my request.

anthony

Hi Anthony, it was me who noticed this, because I played a few Blitz
as Human vs. Zappa and may be I will even do a live Blitz match
against it, as long as Zappa always plays the French exchange vs. me.
I feel I have good chances then to win a 10 games match, when having
always Black

Will Zappa play with this book in RWBC?

Best wishes,
Guenther

The public book doesnt play bad line at all. Zappa?s Books are based on nagged files and pgn files. If Zappa chooses the french exchange mainly (because it is not the only system against the french defense), Zappa statistics consider that exchange has a better overall in the pgn database used to create the book. Even though several moves are nagged with !!, Zappa prefers the move with the better statistics.

Additionally, Anthony requests me to use unusual openings in this opening book because he wanted to see the Zappa?s behavior in such openings, although we had run long tests with such openings.

The public book is optional and if you think, it doesnt fill your expectations, you have other options. Just a matter of tastes.

Regards, Arturo.

Posted: **24 Feb 2005, 19:56**

Hi Anthony, it was me who noticed this, because I played a few Blitz
as Human vs. Zappa and may be I will even do a live Blitz match
against it, as long as Zappa always plays the French exchange vs. me.
I feel I have good chances then to win a 10 games match, when having
always Black Wink

Will Zappa play with this book in RWBC?

Best wishes,
Guenther

Arrow The public book doesnt play bad line at all. Zappa?s Books are based on nagged files and pgn files. If Zappa chooses the french exchange mainly (because it is not the only system against the french defense), Zappa statistics consider that exchange has a better overall in the pgn database used to create the book. Even though several moves are nagged with !!, Zappa prefers the move with the better statistics.

Additionally, Anthony requests me to use unusual openings in this opening book because he wanted to see the Zappa?s behavior in such openings, although we had run long tests with such openings.

The public book is optional and if you think, it doesnt fill your expectations, you have other options. Just a matter of tastes. Very Happy
Regards, Arturo.

Hi Arturo, I have not said the French exchange is a bad line,
but it makes it very easy for Humans to make a draw and it is
one of the most drawish lines around too.
Actually I would always even try to win here for positional
reasons as Black and being a Human.
More important is that it is used as _only_ answer to the French
with 100% probability and it has not the best statistics as the
log from Zappa tells? Did I missunderstand you because it seems
you said something else in your message?

Code: Select all: Probed book successfully (10 entries / 840 bytes / 0 ms). Selected move e4xd5 Nb1-d2 0.0% ?? [ 56.6%, +1988, -1287, =2034] e4-e5 0.0% ?? [ 52.2%, +884, -778, =727] c2-c4 0.0% ?? [ 0.0%, +0, -2, =0] Bc1-e3 0.0% ?? [ 73.5%, +23, -7, =4] e4xd5 100.0% !! [ 48.4%, +395, -444, =718] Bf1-d3 0.0% ?? [ 50.0%, +1, -1, =3] Nb1-c3 0.0% ?? [ 56.3%, +3372, -2301, =2841] c2-c3 0.0% ?? [ 0.0%, +0, -0, =0] f2-f3 0.0% ?? [ 0.0%, +0, -0, =0] f2-f4 0.0% ?? [ 0.0%, +0, -0, =0]

Regards,
Guenther

Winboard Forum

Zappa updated!

Zappa updated!

Re: Zappa updated!

Re: Zappa updated!

Re: Zappa updated!

Re: Zappa updated!

Zappa

Re: Zappa updated!

Re: Zappa updated!

Re: Zappa updated!

Re: Zappa updated!

Re: Zappa updated!

Re: Zappa updated!

Re: Zappa updated!

Re: Zappa updated!

Re: Zappa updated!

Re: Zappa updated!

Re: Zappa updated!

Re: Zappa updated!

Re: Zappa updated!

Re: Zappa updated!