Latest Version

[Tanja1978]

Hello Mike

I would like to thank you for the free upgrade of COW 2016.
Its a very cool tool for learning chess and i will sure improve a little bit in my chess understanding.

I set up my repertoires for black and white. After importing my pgn-files, there is no red T for the Transpositions shown in the header above. But after creating a new EBook and manually adding the lines, the red T is shwon. Is this the normal way how that feature works?
Im also looking for the "Show all move orders"-command. I know it was implemented in earlier Versions. But not in the current?

best regards
Tanja

[Bogg]

Mike:

The reason that I call your limits arbitrary is because neither the compiler nor the computer impose any limits on arithmetic precision. The book "Numerical Recipes in C - The art of Scientific Computing" shows how to do this with numbers stored as strings. You can look at the book at the following url:

http://apps.nrbook.com/c/index.html

20.6 Aritmetic at Arbitrary Precision, pg. 915.

This book has been released in other programming languages if you use something other than 'C'.

You can also find some great information on this subject in the book "Seminumerical Algorithms" v2. You really need to want to learn the subject to wade through this book. To be "fair" to everyone the author made up his own harware with its own assembly language which he uses in all of his example code. But he does show how to perform math on numbers of arbitrary precision. I wrote the addition, subtraction, multiplication and division routines in machine code on a comodore 64. They really are quite simple.

But I mention again that it shouldn't be even that hard:
1) Store the user entered limit in a string.
2) Comparing the stored value to the UCI returned value takes a couple of steps:
a) If the strings are of diferent lengths then the longest is automatically the greatest.
b) if the strings are the same length then compare them.

[Bogg]

Mike:

I was experimenting with IRCs and the ignore blank IRCs feature and it doesn't seem to work. If all of the leaf nodes are blank you assign a White wins or black wins IRC instead of ignoring them and that branch of the tree. Is this the way you expect it to work? If so please explain how to make this of value.

Bogg

[MikeAtBookup]

I don't know if I've ever tried backsolving Informant rate symbols with absolutely none of the leaf nodes tagged with symbols. It should not assign any code if the leaf nodes don't have them.

[Bogg]

Sorry, I again didn't make myself clear. I thought it was obvious that I meant all of the leaf nodes for a branch. That would seem to be the nearly universal case and it doesn't work.

Bogg

[sagisag]

I don't know if I've ever tried backsolving Informant rate symbols with absolutely none of the leaf nodes tagged with symbols. It should not assign any code if the leaf nodes don't have them.

1. Create ebook
2. Import games
3. Backsolve with "Ignore Leaf Nodes with Blank Rates" enabled

and you will see this with IRS.

[MikeAtBookup]

I don't know if I've ever tried backsolving Informant rate symbols with absolutely none of the leaf nodes tagged with symbols. It should not assign any code if the leaf nodes don't have them.

1. Create ebook
2. Import games
3. Backsolve with "Ignore Leaf Nodes with Blank Rates" enabled

and you will see this with IRS.

I'm guessing that you think that importing games has "absolutely none of the leaf nodes tagged with symbols." However, the import process uses the result of the game to set each leaf node. This... is working properly.

[sagisag]

I don't know if I've ever tried backsolving Informant rate symbols with absolutely none of the leaf nodes tagged with symbols. It should not assign any code if the leaf nodes don't have them.

1. Create ebook
2. Import games
3. Backsolve with "Ignore Leaf Nodes with Blank Rates" enabled

and you will see this with IRS.

I'm guessing that you think that importing games has "absolutely none of the leaf nodes tagged with symbols." However, the import process uses the result of the game to set each leaf node. This... is working properly.

Even when the imported games has no result except * (see the 2 games below), COW still added IRS winning for white after backsolving with "Ignore leaf Nodes with Blank Rates" enabled.



The imported games:

Code: Select all

[Event "COW IRS Test"]
[Site "?"]
[Date "????.??.??"]
[Round "?"]
[White "W"]
[Black "B"]
[Result "*"]
[ECO "C84"]
[EventDate "2017.05.17"]

1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.O-O Be7 *

[Event "COW IRS Test"]
[Site "?"]
[Date "????.??.??"]
[Round "?"]
[White "W1"]
[Black "B1"]
[Result "*"]
[ECO "D53"]
[EventDate "2017.05.17"]

1.d4 d5 2.c4 e6 3.Nc3 Nf6 4.Bg5 Be7 *

[MikeAtBookup]

You're right. It's doing something weird. Thanks for that example.

[elirub]

Excuse me is not the last version of COW 2016 build 69, or am I missing something?
Regards elirub