Naum 1.9

[Naum]

New versions everywhere.
Must be the start of the new WBEC cycle

Main changes:

- Finaly found and fixed bug that caused crashes from time to time (this was reported to me by some people almost a year ago)

- Speed improvements. New version is around 25-30% faster then the last version. I still use board[64], but I removed Palm related stuff which enabled me to optimize things a bit.

- There is AMD 64-bit version available. It is not much faster then the 32-bit (around 8-10%).

- Small endgame evaluation improvements. Naum finally understands opposite colored bishops (still not perfect, but better then before).


Regards,
Alex

[Tord Romstad]

Great news, Alex!

After reading a recent post of yours on the WBEC forum, I got the impression that you planned to keep future versions private. I'm happy to see that you have changed your mind, at least for the moment.

Now, if you could just offer us a Linux or OS X version ...

Tord

[Naum]

Hi Tord,

I was so pissed-off by the cloning situation that I considered going private and competing in WBEC only.
I may still do that if I get tired of watching Naum being beaten up by various known and unknown clones.

I also feel pressure to improve Naum, because Glaurung and others are constantly improving. Could you guys stop and give me a break
I don't have time to keep up with you.
By going private, I wouldn't care so much any more.

Regards,
Alex

[Graham Banks]

Thanks for Naum 1.9. I look forward to giving it a run.

I certainly hope you don't go private. You bring many of us a lot of pleasure with your efforts.

Regards, Graham.

[Robert Allgeuer]

Thanks for the release.

A question: which is the default book intdended for use with Naum 1.9 and where can it be obtained?

Thanks
Robert

[Tord Romstad]

Hi Tord,

I was so pissed-off by the cloning situation that I considered going private and competing in WBEC only.

Yes, that's what I recalled.

I may still do that if I get tired of watching Naum being beaten up by various known and unknown clones.

We disagree there, as you know. I don't really think there are a lot of clones in the premier league of WBEC. For all except a handful of the engines, I know enough about the engine or the author to be 99% sure that they are not clones. For the remaining few engines, I don't know enough to feel sure, but I also don't have any direct reason to suspect anything.

Nevertheless, I like your idea that the participants should send their source code for inspection to Dann or some other independent expert.

I also feel pressure to improve Naum, because Glaurung and others are constantly improving. Could you guys stop and give me a break
I don't have time to keep up with you.

I thought I was the one who had difficulties keeping up with the rapid improvement of other engines, including Naum.

Seriously, Glaurung and Naum has followed each other quite closely so far. They emerged on the scene at about the same time, and has been rather close ever since (perhaps with a slight edge for Naum). And based on what you write about the improvements in Naum 1.9, I think that Glaurung 1.0.1 and Naum 1.9 will still be at a very similar level.

I hope that Naum will remain public, and that you will manage to keep your rapid rate of improvement.

Tord

[Naum]

Hi Graham, Robert, Tord,

First of all, I updated the web site with the link to the book. Leo's link got broken when he switched ISPs. Book is the same as the one distributed with the 1.71 version.

As I mentioned, going private would be much easier on my nerves, but I would probably miss watching Naum play in various tournaments. I follow them all, even though I didn't like much watching ProDeo destroying my baby in one of your tourneys Graham

Tord, give me 10 new PCs and Naums rate of improvement will be 10 times faster. This is what frustrates me the most. It is getting very hard to test my changes. I am writting my ideas on paper, saving them to files, writing them on the wall, but in order to test each of them, I need at least 2 days of testing. I have only 1 slow PC dedicated for it and that's not enough.
I can implement around 5 ideas per day, but I need at least 10 days to test each of them. Then you need another 10 days to tune them and test interaction between ideas that passed initial testing.
This means that I have fun 1 day, and then have to wait 20 days to have fun again. Arghhh!

As for the review of the source code, I am still pessimistic it will ever happen. Leo obviously doesn't have time or desire to do something like this, and it may cause some of the authors to withdraw from his tournament, because they won't be willing to provide the source. I like Dan, but isn't he very close to Crafty team and beta-tester for Fruit. He also has his own engine.

It seems that you did a great job with the new Glaurung. I didn't run any matches, but I am affraid that Naum might be falling behind. Anyway, I don't mind being close behind a great developer like you.

Regards,
Alex