DTM data is missing in 6 or 7 syzygy tablebases

[Uri Blass]

For example in KRPPP vs K I can see only DTZ and not DTM and same for other 6 piece tablebases(it seems that there is no problem in the site with the 5 piece tablebases when I see both DTM and DTZ).

https://syzygy-tables.info/?fen=2K5/8/8 ... _w_-_-_0_1

I wonder how much memory you need to add the DTM data at least for the 6 piece tablebases.
I think that the memory that you need to show me all the distance to conversion in the 7 piece tablebases is clearly bigger than the memory that you need to show me the distance to mate in 6 piece tablebases.

[Dann Corbit]

Ronald was working on a DTM version of Syzygy tablebases a while ago.
I do not know if progress has been made.

[jdart]

I am not clear how the website is even getting/outputting DTM info.

The problem with DTM is that with the 50-move rule it is dependent on the game history.

Fathom supports DTM tablebase probing but there are not DTM tablebases to go with that feature. Some previous discussion of DTM here: http://talkchess.com/forum3/viewtopic.php?f=2&t=72158

--Jon

[jp]

I am not clear how the website is even getting/outputting DTM info.

The problem with DTM is that with the 50-move rule it is dependent on the game history.

But "DTM" ignores the 50-move rule, doesn't it?
It'd need to be "DTM50" to include the 50-move rule, and no one's done that so far.

Is the website perhaps giving DTM info from Lomonosov TBs?

[syzygy]

I am not clear how the website is even getting/outputting DTM info.

The problem with DTM is that with the 50-move rule it is dependent on the game history.

But "DTM" ignores the 50-move rule, doesn't it?
It'd need to be "DTM50" to include the 50-move rule, and no one's done that so far.

Is the website perhaps giving DTM info from Lomonosov TBs?

Most likely it is using Gaviota TBs for DTM info:
https://github.com/niklasf/python-chess ... ster/chess