CCRL Discussion Board

From the generation output I saw that you seemed to generate the endings with pawns in them in stages, and saving the output in temporary files, then merging them at the end of the run. Would it be possible, and feasible to expand on that method, and generate e.g. the 6-man pieces in steps of, say,...

Verifying the generated tables is good... Only problem for me is that the package you attached only contains the generator, and the verifyer is a different program if I read the Perl script correctly. Correct, but it is not worth it to run the verifier. Just check the md5sums with fsum and the chec...

The "best permutation" is often not even one of the permutations in the list, and, when it is, it's usually not the best one. It's the same permutation, but encoded differently. I agree it's confusing, but it's how it is ;). I'm hoping this is just an output issue and that it's not select...

It's possible to compile windows executables, see the readme: https://github.com/syzygy1/tb The attached tb.zip file, if it works (maybe executables are blocked, but let's see), contains rtbgen.exe and rtbgenp.exe which should work on CPUs with popcount support. It also contains the run.pl script wh...

If I download a 5 piece tablebase, do I still need to get the 4 piece and the 3 piece tablebases? Or does the 5 piece tablebase include all the positions for 4 and 3 piece tablebase? I'm testing Houdini and Stockfish on my own time and I want to try adding tablebases to the mix as well. For none of...

Thanks! The original position (from the Rybka forum, I think) was: 8/8/8/6kp/2p5/2N2K2/8/3N4 w - - 0 1 Someone complained that my tables did not score this as a win, whereas the Nalimov tables did. Of course the position is a draw by the 50-move rule. What is somewhat funny is that after 1.Ke3, the ...

I'd like to see the DTM50-line for
8/8/8/4k2p/7N/2N2K2/8/8 b - - 0 5

Here's my current draft: http://galen.metapath.org/egtb50/ Some available tablebases attempt to incorporate R50 in various ways, but in my view none are satisfactory. The most common is to add a "DTZ50" metric, which counts down to an irreversible move. It's not quite right, however, as i...

Very interesting results. Funny that the DTM50 record beats the DTM record!

My next step will be to keep only one of the wtm/btm tables in memory during compression and save the other to disk. Pawnful tables use less memory since generation is done per file, but the same trick of saving to disk will be needed to stay under 16 GB. This has now been done (by adding a "-...

This is an interesting development, would certainly like to try if the memory requirements can be reduced to 16G or even 8G :thumbup: 16 GB is doable. I have just changed the indexing scheme during generation of pawnless tables from 10 x 64^5 to 462 x 10^4, so that generation of pawnless tables can...

01) You are using value-only 'WDL-style' EGTs. I had advocated the creation and use of these and indeed, the use of WDL EGTs in the creation of DTx EGTs as this avoids examining potential forced-losses which are in fact draws. I create both WDL and DTZ in one go, so I don't use WDL in the creation ...

I must confess that I did wonder - as the announcement was made on April 1st - whether this was an April Fool's joke. Maybe not the best day for an announcement, but the code is there. :) However, assuming it is not, may I ask if by 'DTZ' you actually mean 'DTZ' - or do you mean 'DTZ50' (where Nali...

Congratulations on this release! Looks like a lot of work went into this. Thanks! It was indeed quite a bit of work. :) I started from the generator I wrote in 2005, but I rewrote practically everything. One of the triggers was the idea of WDL50+ and DTZ50+ that came up some years ago in discussion...

I have just released my tablebase generator for up to 6 pieces on github: https://github.com/syzygy1/tb It generates two sets of files: - WDL files (extension: .rtbw) storing win/draw/loss information for access during search. - DTZ files (extension: .rtbz) storing distance-to-zero information for a...

They are generated and prepared for publication: https://plus.google.com/100454521496393505718/ Does anyone know whether these tablebases will be made publicly available? As I understood it is planned to publish a website where one can query certain positions but does anyone know whether the tableb...

I've not been sitting still, though . Just finished generating the 6-piece WDL50+ and DTZ50+ tables for regular chess Nice. My own vision is that the obvious path for developing new EGTs for regular chess – more men on board – is going to require fast growing computing resources that already exceed...

Syz, good to see you again. This forum is so quiet nowadays… I've not been sitting still, though :). Just finished generating the 6-piece WDL50+ and DTZ50+ tables for regular chess, and I will now generate them for atomic chess. I’ve spotted a small flaw in your line. (...) This effect may be due t...

Amazing stuff… and these “DTM50 new effects” are already visible with only 4 men on board. Except that best play will not lead to MC >= 82 in such positions :wink: My DTZ50 line for this position is: 1. Kb7 Kg7 2. Rf1 Kg6 3. Rh1 Kg7 4. Kc6 Kg6 5. Kd6 Kg7 6. Ke7 Kg6 7. Kf8 Kf5 8. Rxh7 I guess this i...

Maybe clear your browser cache or try with a different browser? I think it simply has to do with the Posts Per Page option of the Rybka Forum. In my case, I have it at 70 . How much is it in your case? (maybe default is 60?) Ok, that explains. I don't have an account on RF, so I guess it's set to w...

See here . OK thanks syzygy! Now I have no problem finding that http://rybkaforum.net/cgi-bin/rybkaforum/topic_show.pl?tid=24816;pg=4 but still cannot find your http://rybkaforum.net/cgi-bin/rybkaforum/topic_show.pl?tid=24816;pg=5 ...simply because page 5 does not exist for me in that Rybka forum t...

ernest wrote:

syzygy wrote:In this file one can find 8/6R1/8/6N1/3k1K2/1B6/7n/7q w - - with mate in 1090 half moves, i.e. 545 full moves.

Thanks!
Do you know where this file comes from?

See here.

See also : http://rybkaforum.net/cgi-bin/rybkaforum/topic_show.pl?tid=24816;pg=5 that page does not seem to exist... The link works for me. kqnkrbn should definitely be the 7-men longest checkmate winner with QN4n1/8/3k4/8/b1K5/8/8/6r1 b - - 0 1 (mate in 545) I have seen it to be a 517 DTC , but do...

Lomonosov supercomputer

Nice hardware

Indeed, I do not include positions with castling rights