Hello to All,
I would like to ask you to give your desired improvements for the next version(s) of Houdini. As Robert Houdart frequently reads the Open-Chess.org forums, he should surely take note of your suggestions. Thanks in advance to All.
I will start myself by giving a few suggestions.
---I suggest that one of the next version of Houdini uses the TripleBases endgame tablebases instead of the current Gaviota tablebases that Houdini 1.5a uses. Why? These TripleBases are stored in RAM and therefore their access is much quicker than when the tablebases are stored on hard disk. On my current humble notebook with a four cores processor, I have 8 GB of RAM and, most of the time, I do not even use 50% of these 8GB of RAM for most of the programs that I use.
I strongly feel that these TripleBases endgame tablebases would strengthen further Houdini.
---I am in favour of the free chess GUIs like Arena and WinBoard. Having said that, I currently use Fritz 12 (the deep version for multiprocessors) and I honestly and fairly think that it is currently the best chess GUI available and by a wide margin.
Therefore, the latest version of Houdini should surely be thoroughly tested first on this particular Fritz 12 GUI. I think Robert Houdart already does that. But it is worth emphasizing this point here.
---In Fritz 12, with the current Houdini 1.5a version when you have installed the Gaviota endgame tablebases, they do not seem to be made available in the Training tab, Endgame Training even though they do work fine with Houdini 1.5a. It would be good if these Gaviota endgame tablebases could also be usable in Training tab, Endgame Training (under Fritz 12).
---During the Elite Match on "TCEC - Elite Match - Live Broadcast" at http://www.tcec-chess.org/tcec_em.php it was interesting to see the various parameters that were recorded while the games were played. Houdini could add several of these parameters to the ones that are already currently available so that one could analyze games played by Houdini even better.
---In Fritz 12, when using Houdini 1.5, for the analysis tools available in Fritz 12, Houdini 1.5 should allow the easy saving of the results obtained when one uses these analysis tools in Fritz 12. I do not see an easy way to save these results in Fritz 12. But I am a recent user of Fritz 12 so I might have missed a few ways to use a few features of Fritz 12 efficiently.
The analysis tools in Fritz 12 that I refer to are in particular:
-Home tab, Infinite Analysis.
-Analysis tab, Full Analysis.
-Analysis tab, Deep Position Analysis.
I will add more suggestions later on.
Thanks to all who will add their own suggestions for the next version(s) of Houdini. Be fully aware that Robert Houdart, the programmer of Houdini, regularly reads the Open-Chess.org forums and therefore I am sure that he will take into account your suggestions given here for the next version(s) of Houdini.
Best Regards to All
MichaelIsGreat
GIVE IMPROVEMENTS THAT YOU WANT IN THE NEXT HOUDINI.
-
MichaelIsGreat
- Posts: 44
- Joined: Thu Dec 30, 2010 3:52 pm
-
MichaelIsGreat
- Posts: 44
- Joined: Thu Dec 30, 2010 3:52 pm
Re: GIVE IMPROVEMENTS THAT YOU WANT IN THE NEXT HOUDINI.
Hello to All,
In my previous post above, one of my suggestions was to ask Robert Houdart, the programmer of Houdini, to improve dramatically the analysis capability of Houdini under Fritz 12. Especially the ability to be able to save for review the analysis that has been done by Houdini under the various analysis tool available in Fritz 12, in particular the analysis tools available in Fritz 12 that I have indicated clearly. I also suggested added several parameters, some of these new parameters could be directly linked to analysis.
I have just started to read a post on this matter of improving analysis capabilities for Stockfish. As Stockfish is an open source program, the programmer of Houdini could have a look at the ways proposed in this post to improve Stockfish analysis capabilities. This post is "Designing an analysis friendly Stockfish?" at http://www.open-chess.org/viewtopic.php?f=5&t=1042
According to this post on Stockfish, it would seem that, in order for a chess engine to be able to analyse well, it has to be coded by the programmer of the particular chess engine targeted for improvement to analyze games. I am sure Robert will have a look at this particular post and that he will assess the possibilities on this matter of improving further the analysis capabilities of Houdini in Fritz 12 for a next version in the future.
Another idea that comes to mind. If to add code for improving the analysis capabilities of Houdini in Fritz 12 would be detrimental to the strength of Houdini, I suggest that the analysis capabilities of Houdini be implemented by Robert in such a way that they would be easily enabled or disabled by the user of Houdini in Fritz 12, depending on the goal that the user has when using Houdini. For instance, through a few parameters options for Houdini in Fritz 12. This way, the user would be able to use Houdini for two different purposes: for pure strength or for analysis. Just in case the improved analysis capabilities implemented in Houdini would be detrimental to the strength of Houdini, something Robert would have to check I assume.
A chess engine program is not only a great adversary against other chess engine programs and against the strongest humans (even though chess engine programs are now simply too strong for any human to compete), a chess engine program is also used by many mainly as an analysis tool. That is the reason why to improve further Houdini capabilities in this particular area of analysis in Fritz 12 is very important I believe, as Houdini is currently significantly better than the best second chess engine program in the world.
Thanks to Robert Houdart for having given us the best chess engine program in the world so far, Houdini 1.5a. He did an amazing job with Houdini. And I am sure that there are opportunities to make Houdini even stronger in the future despite the fact that its current strength has reached the highest ELOs that any chess engine program has ever reached so far.
Regards to All
MichaelIsGreat
In my previous post above, one of my suggestions was to ask Robert Houdart, the programmer of Houdini, to improve dramatically the analysis capability of Houdini under Fritz 12. Especially the ability to be able to save for review the analysis that has been done by Houdini under the various analysis tool available in Fritz 12, in particular the analysis tools available in Fritz 12 that I have indicated clearly. I also suggested added several parameters, some of these new parameters could be directly linked to analysis.
I have just started to read a post on this matter of improving analysis capabilities for Stockfish. As Stockfish is an open source program, the programmer of Houdini could have a look at the ways proposed in this post to improve Stockfish analysis capabilities. This post is "Designing an analysis friendly Stockfish?" at http://www.open-chess.org/viewtopic.php?f=5&t=1042
According to this post on Stockfish, it would seem that, in order for a chess engine to be able to analyse well, it has to be coded by the programmer of the particular chess engine targeted for improvement to analyze games. I am sure Robert will have a look at this particular post and that he will assess the possibilities on this matter of improving further the analysis capabilities of Houdini in Fritz 12 for a next version in the future.
Another idea that comes to mind. If to add code for improving the analysis capabilities of Houdini in Fritz 12 would be detrimental to the strength of Houdini, I suggest that the analysis capabilities of Houdini be implemented by Robert in such a way that they would be easily enabled or disabled by the user of Houdini in Fritz 12, depending on the goal that the user has when using Houdini. For instance, through a few parameters options for Houdini in Fritz 12. This way, the user would be able to use Houdini for two different purposes: for pure strength or for analysis. Just in case the improved analysis capabilities implemented in Houdini would be detrimental to the strength of Houdini, something Robert would have to check I assume.
A chess engine program is not only a great adversary against other chess engine programs and against the strongest humans (even though chess engine programs are now simply too strong for any human to compete), a chess engine program is also used by many mainly as an analysis tool. That is the reason why to improve further Houdini capabilities in this particular area of analysis in Fritz 12 is very important I believe, as Houdini is currently significantly better than the best second chess engine program in the world.
Thanks to Robert Houdart for having given us the best chess engine program in the world so far, Houdini 1.5a. He did an amazing job with Houdini. And I am sure that there are opportunities to make Houdini even stronger in the future despite the fact that its current strength has reached the highest ELOs that any chess engine program has ever reached so far.
Regards to All
MichaelIsGreat