- Rybka 4 will be a normal UCI engine, without copy protection, and will be available in the first part of May.
- There will be separate single-processor and multi-processor versions.
- Full chess analysis packages which include Rybka 4 will be made by ChessBase (www.chessbase.com) and Convekta/ChessOK (www.chessok.com).
- Plain Rybka 4 UCI without GUI for download only will be available from RybkaChess (www.rybkachess.com).
- All of these versions of Rybka 4 will be identical and can be used in any UCI-compliant GUI.
- The Rybka 4 book by Jiří Dufek will be available as a separate item from both ChessBase and Convekta. The two versions of the book will be 'essentially identical', although in different formats.
- Rybka 4 book compatibility will work as follows:
* ChessBase software users will require the ChessBase version of the book.
* Aquarium users can use either version of the book.
* Other software users are encouraged to upgrade their chess software. :)
I'll do my best to answer all questions after the release.
Vas
Now, let's start with some questions.
1, Is the ELO improvement still top secret ? ;-)
1a, If so, is it because you are still tuning the program, (especially against the new and foreseable competitors)?
2, Did you improve somewhat the scaling factor for 2, 4, 8... cores relatively from Rybka 3?
3, Are the Rybka 3+ bugfix (bishop under promotion, hashtable bug, etc) incorporated in Rybka 4?
4, Could you say anything about possible playing style change, program personnalities, analysis enhancements, possible new UCI options?
5, What about those (hateful) rumors of a dumbed down version to preserve some elbow room for the rental model?
6, Given the fresh reception to the rental model, is that project still on track?
7, Are you still 100% involved with Rybka or is your disappearance from this forum connected with the pursuit of other interests?
Thanks for any clarification and information sharing.
I am also not sure if Vas know the final version that he is going to release and he may find a productive change earlier to the release date so the program may be stronger than what he has now.
I disagree.
If he find a productive change earlier to the release date, he will use it for the Rental Rybka.
Vas can give an estimate, and he can give the conditions of his test, so we can evaluate it correctly.
It is better for buyers to have an idea about strength of the product that they pay for.
> 1, Is the ELO improvement still top secret ?
Performance is best measured by independent testers.
> 1a, If so, is it because you are still tuning the program, (especially against the new and foreseable competitors)?
> 2, Did you improve somewhat the scaling factor for 2, 4, 8... cores relatively from Rybka 3?
Not substantially.
> 3, Are the Rybka 3+ bugfix (bishop under promotion, hashtable bug, etc) incorporated in Rybka 4?
I'll go through these items after the release. Some are fixed, some are not.
> 4, Could you say anything about possible playing style change, program personnalities, analysis enhancements, possible new UCI options?
There will be some new things, I'll document them after the release.
> 5, What about those (hateful) rumors of a dumbed done version to preserve some elbow room for the rental model?
> 6, Given the fresh reception to the rental model, is that project still on track?
Yes. The reception so far has been good. However, the public launch is still some time away (months, not weeks). More info will be available later.
> 7, Are you still 100% involved with Rybka or is your disappearance from this forum connected with the pursuit of other interests?
Yes, I am working only on Rybka.
Vas
Anyway, thanks for such a quick answer, keep the good work!
Cheers
I wonder if you could put in the same package many different versions (even a little bit weaker) of rybka4 engines, like you did for rybka3 with FRC, Human etc. It's very important for the chess community to have the feeling that we did not pay a lot for an engine (if you release Rybka4, Rybka4 FRC, Rybka4 Agressive, Rybka4 Strategic, Rybka4 Human for example with a price of 50 Euros for single CPU and 100 Euros for Multiple CPU (and may be Cluster Rybka4 in another package, we will be not only very happy to test all these engines, this will make the community speek much more about your product and this will also make free advertising for you.
> - Rybka 4 will be a normal UCI engine, without copy protection, and will be available in the first part of May.
First out of how many parts?
I'm looking forward to buying this!
Will you do some limited beta testing before the May release and if so can I apply?
It's very important for the chess community to have the feeling that we did not pay a lot for an engine.
Releasing some beta engines is as always wellcomed by all of us, and it gives us the feeling that for the same 50 or 100 bucks we pay for much more than a single engine. Example of possible names (Rybka4, Rybka4 FRC, Rybka4 Agressive, Rybka4 Strategic, Rybka4 Human, but you can also try to make Kasparov Rybka Tal Rybka (different personnalities)), and why not an extra CLUSTER Rybka for those who are ready to pay more.
Such a strategy will also provide you a free advertising as all the curious chess community will try to make different matches to test your different engines. As you said before there is a 3000 different single rybka versions, with different strengths and style of play, it seems that you can choose aming those beta versions (may be try to improve them a lot by taking some of your latest best ideas that you use for your final product).
You can also give games for the community with names like Rybka 3.xxxxx to analyse and to help you choose the name but also the style of play of the engines.
I'm ready to help for thsi task for free (mail: dr_zied_haddad at yahoo dot fr.
I got a friend with which i discussed a lot a few years ago, and that was involved in the development of Rybka2 versions. I gave him a lot of ideas (when i was used to develop settings for Shredder) which i think you are using right now; I'm happy to see that some of my ideas are used by others and are working. These ideas are not so clever, but practicle, and i'm sure that without the help of this friend you ae able to do the same work or even better.
Hope to see a good product with a good strategy, without having the chess community being suspicous about you making money by giving a beta product.
What we must be sure about is that even the final version is still a beta.
Your strategy to release a dozen of different engines as it was for Rybka 1 1.2 2 etc was good, the only problem is that you released so much beta versions in a very short period of time, so think about releasing a beta each 21-30 days after working on developing these versions Think also about enlarging your circle of persons of Confidence with some fresh ideas and without any interest. Some players that use your engine for example to analyse correspondance games and play correspondance games, but also to make some tournament tests.
It's always interesting to see how your engine does in different situations.
Sincerely yours
Zied HADDAD, M.D, Ph.D in anaesthesiology and intensive care.
They can only "read" the chessbase version, to clarify :)
Besides as you said in your interview in NewInChess, Engines are tools and we should treat them that way.
It's very important for the chess community to have the feeling that we did not pay a lot for an engine.
Releasing some beta engines is as always wellcomed by all of us, and it gives us the feeling that for the same 50 or 100 bucks we pay for much more than a single engine. Example of possible names (Rybka4, Rybka4 FRC, Rybka4 Agressive, Rybka4 Strategic, Rybka4 Human, but you can also try to make Kasparov Rybka Tal Rybka (different personnalities)), and why not an extra CLUSTER Rybka for those who are ready to pay more.
Such a strategy will also provide you a free advertising as all the curious chess community will try to make different matches to test your different engines. As you said before there is a 3000 different single rybka versions, with different strengths and style of play, it seems that you can choose aming those beta versions (may be try to improve them a lot by taking some of your latest best ideas that you use for your final product).
You can also give games for the community with names like Rybka 3.xxxxx to analyse and to help you choose the name but also the style of play of the engines.
I'm ready to help for this task for free (mail: dr_zied_haddad at yahoo dot fr.
Your strategy to release a dozen of different engines as it was for Rybka 1 1.2 2 etc was good, the only problem is that you released so much beta versions in a very short period of time, so think about releasing a beta each 21-30 days after working on developing these versions Think also about enlarging your circle of persons of Confidence with some fresh ideas and without any interest. Some players that use your engine for example to analyse correspondance games and play correspondance games, but also to make some tournament tests.
It's always interesting to see how your engine does in different situations.
Finally, you must be honest and give more informations about the strength of Rybka4. Poeple will buy it even if it's only 25 rating points more (depends on the number of engines and price). The addicted to computer chess will buy your engines, the others will wait for the first published results of different tests to assess it's strength. With a strategy not revealing the real strength of Rybka4, your work will not be granted success.
With many free engines (no matter the controversy concerning stolen rybka code etc.) that are stronger than Rybka3, few are poeple who'd like to spend a few bucks for a product that doesn't tell the truth. But you'll get allies even if you are weaker, but honest.
Sincerely yours
Zied HADDAD, M.D, Ph.D in anaesthesiology and intensive care.
Gordon B.Sc, MSc, PhD (Theoretical Physics)
Norris, White Belt, Black Belt, Superman (Defender of all that is Holy)
Paulo Soares
www.chessok.com with Aquarium interface
www.rybkachess.com without interface
It's not available yet.
The running announcements of a mom-and-pop store shouldn't taken seriously. ;-)
Vas should take a lesson from Apple and maintain secrecy until he is ready make a big announcement on his own site and on his own terms rather than letting others dribble out his message. No marketeer is he.
Paulo Soares
> * Other software users are encouraged to upgrade their chess software. <img class="sic sic_sml_pos" src="/mwf/epx.png" alt=":-)">
Now why on earth would anyone downgrade their software to Chessbase or Aquarium?
Such bloated software is only required by those who like "pretty" (pretty rubbish to be precise).
:)
Christopher
Paulo Soares
> Because users like?
Everything is relative. White wall tyres on a tractor does not really cut it, unless you want to look good in the middle of a field, in the middle of nowhere.
Only reason why it is in Chessbase is due to distribution. It is not for computer chess reasons.
It has nothing to do with what users like. It has to do with what they see in shops.
become blind to analyze the Chessbase products. But what´s the problem with
Aquarium?
I will sleep,5:00 am here.A good day for you.
Paulo Soares
So probably we are missing out.
Eelco
Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill
