Houdini Versions

[JohnS]

"Houdini 1.03 is on schedule for July 15, I'll consider your request for the version after 1.03."

Can you tell us what you have planned for version 1.03? Thanks.

[Robert Houdart]

@Royb,
I will first release Houdini 1.03 for Windows, then update the code base for Linux and hopefully manage to come out with a competitive Linux compile. I haven't touched Linux since 2004, curious to see how it's evolved over the last 6 years.

@JohnS,
Houdini 1.03 will include multi-PV and searchmoves, large page support, and also some improvement in strength.

Cheers,
Robert

[benstoker]

@Royb,
I will first release Houdini 1.03 for Windows, then update the code base for Linux and hopefully manage to come out with a competitive Linux compile. I haven't touched Linux since 2004, curious to see how it's evolved over the last 6 years.

@JohnS,
Houdini 1.03 will include multi-PV and searchmoves, large page support, and also some improvement in strength.

Cheers,
Robert

I suggest Ubuntu 10.04 (Although the new brownish colour scheme/theme is a bit odd IMO -- but, that's irrelevant)

Doing a Linux compile will be VERY much appreciated. For one thing, the ease of simply running the engine on a remote machine with a simple command like 'ssh remote.machine.com houdini_linux' will be wonderful. And then another engine can run on the local machine.

[benstoker]

@Royb,
I will first release Houdini 1.03 for Windows, then update the code base for Linux and hopefully manage to come out with a competitive Linux compile. I haven't touched Linux since 2004, curious to see how it's evolved over the last 6 years.

@JohnS,
Houdini 1.03 will include multi-PV and searchmoves, large page support, and also some improvement in strength.

Cheers,
Robert

BTW, if I may butt in again, I want to make a pitch again to you to 'go open source.' I think you said earlier you have no plans of going commercial. Since that's the case, take notice of how the Stockfish team develops. Because it's open source, people can and do submit patches to the core developers and there are a lot of folk who contrib to that project. Same goes for Hyatt's crafty. Being based on IPPO*/Ivanhoe, I would wager you could benefit and the Houdini project could benefit. Who cares if the codebase is a tangled web of code? Perhaps with your leadership, others, I think would naturally jump in. The interest is there because it's an exceptionally strong engine. I think the Windows development world has understandable difficulty with the "strange" open source idea. But, I hope my sales pitch here is a little food for your thought. Also, I'll add that the recognition of contribution will be yours; I mean, I would not think your recognition would diminish as a result of going open source and thus opening up contributions to your project.

Ok. I'll stop now.

[benstoker]

@Royb,
I will first release Houdini 1.03 for Windows, then update the code base for Linux and hopefully manage to come out with a competitive Linux compile. I haven't touched Linux since 2004, curious to see how it's evolved over the last 6 years.

@JohnS,
Houdini 1.03 will include multi-PV and searchmoves, large page support, and also some improvement in strength.

Cheers,
Robert

BTW, if I may butt in again, I want to make a pitch again to you to 'go open source.' I think you said earlier you have no plans of going commercial. Since that's the case, take notice of how the Stockfish team develops. Because it's open source, people can and do submit patches to the core developers and there are a lot of folk who contrib to that project. Same goes for Hyatt's crafty. Being based on IPPO*/Ivanhoe, I would wager you could benefit and the Houdini project could benefit. Who cares if the codebase is a tangled web of code? Perhaps with your leadership, others, I think would naturally jump in. The interest is there because it's an exceptionally strong engine. I think the Windows development world has understandable difficulty with the "strange" open source idea. But, I hope my sales pitch here is a little food for your thought. Also, I'll add that the recognition of contribution will be yours; I mean, I would not think your recognition would diminish as a result of going open source and thus opening up the possibility of contributions from third parties to your project.

Ok. I'll stop now.

[Martin Thoresen]

@Royb,
I will first release Houdini 1.03 for Windows, then update the code base for Linux and hopefully manage to come out with a competitive Linux compile. I haven't touched Linux since 2004, curious to see how it's evolved over the last 6 years.

@JohnS,
Houdini 1.03 will include multi-PV and searchmoves, large page support, and also some improvement in strength.

Cheers,
Robert

Robert, I am interested in if you are following the Houdini long time-control match I have going (soon finished).

http://home.halden.net/mordor/match/broadcast.php

And your thoughts surrounding it.

[Robert Houdart]

Robert, I am interested in if you are following the Houdini long time-control match I have going (soon finished).

http://home.halden.net/mordor/match/broadcast.php

And your thoughts surrounding it.

Martin,

I'm a regular follower of this high-level match.
Some thoughts:
- It's been quite an exciting match, games have been more entertaining than I expected. There's been some extremely good, but also some disappointing play by both engines.
- In the first half of the tournament Lady Fortune was on Rybka's side, in the second half the tables have been turned. The current score line of 23-20 appears to be a fair reflection of the match so far.
- Based on this match only I would conclude that Rybka 4 appears to be somewhat stronger in the opening and early midgame, whereas Houdini seems to have a slight edge in the endgame. Of course it's dangerous to jump to conclusions based on just 40 games, the random factor is very high.
- When the match is over I will certainly study some of the games more in detail , there are things to be learned that will help to make Houdini stronger.

Thanks for creating this event!

Robert

[kingliveson]

After finding some IvanHoe settings that's outperforming Rybka 4, am a little curious as to how Houdini fairs against IvanHoe 9.55b. The problem testing Houdini on my AMD systems is that, Houdini I believe has Intel specific optimization. There was a thread on how Houdini's performance lags on AMD architecture -- cant find it now. One with Intel system will have to do the testing. Perhaps Mr. Houdart is a Capitalist working on behalf of Intel...

[AnthonyTheSage]

The release of Houdini 1.03 is going to be a monumentus occasion. Multi Pv and the Works, what a treat. I apologize to all you Ivan fans out there but I think Houdini is more of a players engine. Ivan seems more suited for programmers who like to tweak and fiddle and compile. I mean you never know what your gonna get from version to version. Theres so many tweaks to the playing style that I bet they analyze nothing like they play. So cheers to Robert, keep it fast and keep it simple.

[jeandis]

Hi Robert,

I used your Houdini 1.03a engine with my own book Mapanare 1.0b (I am the author of Mapanare) with a lot of good results. I used split_depth 13 instead of 10. My hardware is an Intel quad 6600 at 2.4 ghz with 4 gigas of memories RAM and using the windowz 7 ultimate. To increase the results what it'd be the best split_depth, 10, 11 or 12?
Hoping news from you