Author: Steven Edwards
Date: 13:38:53 03/15/04
Go up one level in this thread
On March 15, 2004 at 14:29:58, Robert Hyatt wrote: >On March 15, 2004 at 14:00:15, Anthony Cozzie wrote: >>>-> the bottom line: >>>perhaps you are onto something great, we will see. until then, i (and i believe >>>many others here) prefer to think you are somewhere between slightly and stark >>>raving mad with your project goals. AFAIK you have absolutely nothing to show up >>>to now (lots of code, but i'd like to see some chess moves...). AFAIK your >>>engine has yet to make it's first move. it is a mystery to me how you can first >>>implement opening book and tablebase access before addressing the real problems >>>in your approach. it is also a mystery to me how you can spend lots of time >>>posting here about what your program will be able to do in the end, when it >>>can't do anything right now. 1. As Symbolic is to be a real, competing program then it has to have tablebases, an opening library, and other niceties that are unneeded in a paper program. These were included early to both test the toolkit interface and also because the top level portion of Symbolic's move selector (now complete), needed these to detect whether or not a full planning search is needed. 2. As a result of the above integration and other completed code, it is now much easier to test Symbolic. In fact, as the toolkit provides an interactive command processor along with automated EPD test suite sequencing, Symbolic has not only been able to make a move, but to run a complete suite like BWTC all by itself. Of course, at this point Symbolic can only solve mate in one and tablebase positions, so nothing really exciting is happening. Yet. >>>good luck - you will need it... >>I think Symbolic is an interesting project, simply because it is different. I spent a lot of time thinking about other approaches. One alternative is to try a non-planning "whole tree at once" method like Berliner's B* search. As Symbolic already supports keeping the whole, annotated tree in memory, I could get a version of B* running in a day - if only someone could supply me with a pair of Berliner's magical bound estimating functions. But B* and similar schemes do not model human thought so even if successful they are not all that interesting to me. >>Whether it works or (probably) doesn't work, it provides a new datapoint, and >>perhaps some new perspective. However, Steven's *plans* for symbolic are not >>nearly as interesting as his *results*, and the former has been much more >>forthcoming than the latter. Hmmn. Maybe I should offer a wager or two here to the doubters. Like, if I can't get this to work, then I'll stop complaining about the mundane nature of traditional A/B searchers; if I do get it too work, each doubter can send me a new battery for one of my Macintosh notebooks. (Approx. US$150 each.) >Well, the good news is that Steven's last name is not Botvinnik. :) I suspect >that he will eventually produce something, unlike the "Pioneer" project that >produced reams of promises and faked results... :) I still don't know if Mikhail Botvinnik has being intentionally dishonest or merely self-deluded. I think that he was able to get by for so long under the old USSR (because of politics) and so was unprepared to defend Pioneer to Western researchers uncowed by his Kremlin connections. The real losers here were the sincere Soviet chess programmers whose work lost support and recognition that undeservedly went to Botvinnik. (I've always thought that Mikhail Tal was the best player ever produced by the USSR and if he hadn't have had various health problems along with a bit less boozing, perhaps Botvinnik never would have held the WC title.) Also: Since I trust you Bob, I'm willing to send you a gzip tar file of Symbolic's entire source under non-disclosure once a month or so and you can give your opinion to CCC as to whether or not it's a real program. Then the doubters can give a hard time to you instead of to me. Heh, heh.
This page took 0.01 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.