Texel 1.02

[petero2]

Version 1.02 of my chess engine Texel is now available for download from http://dl.dropboxusercontent.com/u/89684995/texel102.7z

There are 4 compiled versions included:

texel-arm : For the armv7-a architecture. Should work on most modern android devices.
texel64 : For 64-bit linux system with SSE42 and POPCOUNT support.
texel64.exe : For 64-bit windows systems with SSE42 and POPCOUNT support.
texel64old.exe : For 64-bit windows systems without SSE42 and POPCOUNT.

The windows versions are about 10% slower than the corresponding linux versions, see http://web.comhem.se/petero2home/javach ... html#texel. Perhaps some windows compiler guru can improve on that situation.

Texel 1.02 scores about 64% (99 elo) against Texel 1.01 on my linux computer at very fast time controls, but I'm sure the difference is smaller at longer time controls against different engines.

There have been a number of mostly small changes since version 1.01:

- Fixed KNNpK evaluation bug.
- Fixed infinite loop when engine received EOF on standard input.
- Only do one check in q-search.
- Speeded up and improved the search tree log file analyzer.
- Remember history information between searches.
- Remember best move (for LMR threat detection) also when there is a transposition table hit or when going into q-search on the next ply.
- Cache line aligned the main transposition table and the pawn hash table.
- Use cache-optimized magic bitboard constants.
- Handle stalemate in kqkp endings with pawn on a2/h2 and king on a1/h1.
- Detect illegal moves before playing them. Makes code look nicer.
- Reduce aspiration window from 20cp to 15cp.
- Gradually widen aspiration window after fail high/low at root.
- Store stale-mate scores in transposition table.
- Don't do LMP if best move so far is negative mate score.
- Smaller razor margin at depth <= 1.
- Support for SSE4.2 and POPCOUNT.
- Added knight mobility evaluation.
- Added KRPKR end game evaluation based on table base statistics.
- Use different bishop piece square table in the end game.
- Some speed optimizations.
- Fixed castle bonus asymmetry.
- Use an incrementally updated material identifier to simplify end game classification.
- When ordering root moves, use accumulated node counts instead of just last ply counts.
- Added evaluation term for bad bishops.
- Rescale history counters before starting a search, so that they adapt faster to the new situation, but still remember information from the previous search.
- Use a material hash table with material score and interpolation factors for other evaluation terms.
- Added evaluation term for redundancy of major pieces.
- Made history aging symmetric with respect to success/failure.
- Implemented knight outpost evaluation.
- Optimized pawn structure, major piece redundancy and knight outpost weights using CLOP.
- Disabled late move pruning in late end games.
- Fixed crash if EOF is received on standard input while engine is thinking.
- Better time management in "ponder on" games with increment.
- When only one legal move is available, think slightly longer than 2 ply.

As always the talkchess forum and the chess programming wiki have been great sources of inspiration for new ideas to test.

[Werner]

Thanks Peter,
nice to see you are improving your engine!

Have a nice Weekend!

Werner
CEGT

[Chan Rasjid]

The last time not too long ago (3/4 months), the linux texel version that I got played strange. It probably had a bug introduced; played very fast and losing most of the time.

Rasjid.

[petero2]

The last time not too long ago (3/4 months), the linux texel version that I got played strange. It probably had a bug introduced; played very fast and losing most of the time.

Which version did you get? Did you compile it yourself from source code? Do you have any saved games with time usage information?

Also it is worth noting that texel tries to keep at least 1 second remaining on its clock, so if you played hyper-bullet games where the total game time is not much larger than 1 second, texel may play a lot weaker than programs that have more aggressive time usage in hyper-bullet games.

Another thing is that 1.01 played too fast in ponder-on games with increment. This is fixed in version 1.02.

[Graham Banks]

Version 1.02 of my chess engine Texel is now available for download from http://dl.dropboxusercontent.com/u/89684995/texel102.7z.........

Thanks Peter.

[Chan Rasjid]

The last time not too long ago (3/4 months), the linux texel version that I got played strange. It probably had a bug introduced; played very fast and losing most of the time.

Which version did you get? Did you compile it yourself from source code? Do you have any saved games with time usage information?

Also it is worth noting that texel tries to keep at least 1 second remaining on its clock, so if you played hyper-bullet games where the total game time is not much larger than 1 second, texel may play a lot weaker than programs that have more aggressive time usage in hyper-bullet games.

Another thing is that 1.01 played too fast in ponder-on games with increment. This is fixed in version 1.02.

I have deleted the executable and there is no info available.

I think I was not mistaken as I knew an earlier version was very much stronger than mine, probably at the class of fruit.

I play only games of 60moves in 30secs with same time control added.

Rasjid.

[petero2]

The last time not too long ago (3/4 months), the linux texel version that I got played strange. It probably had a bug introduced; played very fast and losing most of the time.

Which version did you get? Did you compile it yourself from source code? Do you have any saved games with time usage information?

Also it is worth noting that texel tries to keep at least 1 second remaining on its clock, so if you played hyper-bullet games where the total game time is not much larger than 1 second, texel may play a lot weaker than programs that have more aggressive time usage in hyper-bullet games.

Another thing is that 1.01 played too fast in ponder-on games with increment. This is fixed in version 1.02.

I have deleted the executable and there is no info available.

I think I was not mistaken as I knew an earlier version was very much stronger than mine, probably at the class of fruit.

I play only games of 60moves in 30secs with same time control added.

I have only released two versions of texel, 1.01 over a year ago, and 1.02 today, so I don't know what this version you are referring to is. Do you know where you got if from?

[Werner]

JA compiles are online.
Thanks Jim!
Werner

[Sylwy]

THANK YOU !

Texel 1.0.1 in my "10+1 w32 Tournament":

http://talkchess.com/forum/viewtopic.ph ... 22&t=48225

Regards,
SilvianR

[Chan Rasjid]

The last time not too long ago (3/4 months), the linux texel version that I got played strange. It probably had a bug introduced; played very fast and losing most of the time.

Which version did you get? Did you compile it yourself from source code? Do you have any saved games with time usage information?

Also it is worth noting that texel tries to keep at least 1 second remaining on its clock, so if you played hyper-bullet games where the total game time is not much larger than 1 second, texel may play a lot weaker than programs that have more aggressive time usage in hyper-bullet games.

Another thing is that 1.01 played too fast in ponder-on games with increment. This is fixed in version 1.02.

I have deleted the executable and there is no info available.

I think I was not mistaken as I knew an earlier version was very much stronger than mine, probably at the class of fruit.

I play only games of 60moves in 30secs with same time control added.

I have only released two versions of texel, 1.01 over a year ago, and 1.02 today, so I don't know what this version you are referring to is. Do you know where you got if from?

I don't really know. If you only had one earlier version before this, it is surprising.

I remember something like getting an update, usually from Jim Ablets site or maybe a link from others to your homepage. It was a precompiled linux version. Then it was of no use as it loses all the time, unlike before - the same vesion as you seem to say.

Rasjid.