Test suite bug

[GoogleCodeExporter]

What steps will reproduce the problem?
1. bin/gray
2. test path_to/ECE3.PGN

What is the expected output? What do you see instead?
Test suite finishes with 99% correct solutions.
Instead, GM resigned at position 413. And if GM resigns a
position, the test doesn't continue.

Please use labels and text to provide additional information.
To fix:
- GM should continue after a resign
- I think GM might be running out of time. After each test position
  we might have to reset time.

Original issue reported on code.google.com by jonne.z...@gmail.com on 28 Dec 2007 at 8:43

[GoogleCodeExporter]

Fixed in r1412, going to test it now.

Original comment by jonne.z...@gmail.com on 27 Jan 2008 at 6:38

[GoogleCodeExporter]

It seems I reset the clock correctly. But there's still another problem.

...
../TestSuites/Tactical/ECE3.PGN 568 : 'Ke5' != 'f5'
Clock: 40<40 3+0
resign

Here Gray matter resigns at position 569, but this seems undeterministic. 
Running
again it either doesn't resign, or sometimes at another (random) position.

Why? The clock says play the next 40 moves in 3 csecs.
I also tried with 30 csecs, also got the error sometimes.

Does this mean GM cannot even find a legal move within time?
And then resigns? Hmmzzz

Original comment by jonne.z...@gmail.com on 27 Jan 2008 at 7:12

[GoogleCodeExporter]

I know it's the last line of 
  void search_mtdf::iterate(int s)

that calls print_result with
a null move.

Maybe the status is changing too quickly, we had a problem like that before...

Original comment by jonne.z...@gmail.com on 27 Jan 2008 at 7:26

[GoogleCodeExporter]

Could someone tell me where to find this ECE3.PGN file, so i can fix this when 
I get
time?  Also, perhaps we could put all of the other test files you all may have 
on the
downloads, so others can mess with them?

Mike

Original comment by protonsp...@gmail.com on 10 Jul 2009 at 9:31

[GoogleCodeExporter]

Sorry Mike, I have problems finding it myself (my laptop crashed some time ago, 
but I
might still have it at some other place). I'm not sure whether we're allowed to 
put a
copy on our project site. If so, we should certainly do that.

Here are some other interesting links:
- http://www.arasanchess.org/tests.html

ECE stands for "Encyclopedia of Chess Endings" if I'm correct. I thought it 
would be
better to test Gray Matter on endgames first, just like I think a human player 
should
start by studying endgame positions.

Aha, I think I've found them all here ->
http://code.google.com/p/tropo/source/detail?r=33

Jonne.

Original comment by jonne.z...@gmail.com on 4 Aug 2009 at 1:19

[GoogleCodeExporter]

This also seems very suitable as a test suite:
http://www.jakob.at/steffen/hossa/testsuites/wac.epd

Original comment by jonne.z...@gmail.com on 3 Sep 2009 at 7:44

[GoogleCodeExporter]

I've successfully done the test many times.  There do not seem to be any issues 
in the current version.

Original comment by protonsp...@gmail.com on 29 Jul 2010 at 9:45

• Changed state: Done