Website founded by
Milan Velimirović
in 2006

22:49 UTC
ISC 2021



Remember me

Forgot your
Click here!
to create your account if you don't already have one.

Rating lists


MatPlus.Net Forum Internet and Computing Natch as Natch can
You can only view this page!
(1) Posted by Hauke Reddmann [Wednesday, Jan 6, 2021 19:11]

Natch as Natch can

For what reason'o'ever, I can't redirect the output of Natch
into a file. I wrote natch.bat containing
natch.exe <example.txt >out.txt
but the output file stays empty. Neither does >> or |more help.
(Powershell: hangs up with that batch)

(Read Only)pid=20390
(2) Posted by Joost de Heer [Wednesday, Jan 6, 2021 20:59]

natch -o output

Hint: What switch do you usually use for help?
(Read Only)pid=20391
(3) Posted by Andrew Buchanan [Friday, Jan 8, 2021 17:06]

I normally run Natch from a PC command prompt. In the directory, I've renamed one file as "Natch -t 32 -l e -v -o outputn.txt inputn.txt" so I just cd to the directory, type dir, then copy/paste the file name to run Natch. I no longer know what all the letters in the incantation mean.

I'm still surprised that Pascal Wassong is ok that a position which is illegal due to wrong side checking nevertheless can have solutions in Natch! I thought that the *only* thing you know about a general illegal position is that it has zero proof games. Other PG engines spot this. But some people seem to panic about wrong side checking. It's only illegal, nothing more: and curiously the rules of chess are such that one can just play forwards. And obv the rules must still work in illegal positions, or everyone would have to do retro analysis before they can play any move in a game where they haven't got the game score.

But Natch is pretty awesome though: it's very fast and although the number of solutions is just indicative, if you want an accurate count it can be used in combination with one of the slower engines that counts each solution properly. Stick all the Natch solutions into a spreadsheet, and decompose them into constituent moves. This indicates what are all the legal moves: Natch's uncertainty is only about the sequences. So can retract last moves in Jacobi or Popeye to get shorter PGs which they can probably solve, or in Jacobi can put in constraints. This multi-engine trick is very useful. And can do it also before Natch finishes.

Does Euclide still stop as soon as it finds what I think is the first dual? That is a showstopper for me: a big clue in PG exploration is the rough number of solutions which are appearing, and the timing of their arrival. Obv I terminate the run after hundreds, but Euclide stopped at just 4, which made it unusable for composition.
(Read Only)pid=20403
(4) Posted by Hauke Reddmann [Friday, Jan 8, 2021 22:47]

Well, as a CS student I say: It's better to natch than to not natch,
it's better to have a RTFM than to have not (since none of the files
had a text file extension, I guess its "readme", but I won't be
commanded around by files :-) but best is the program which doesn't
need a RTFM in the first place. (The RTFM of *my* proggies for my
running CS thesis will take about the same size as the thesis itself,
pot kettle, lasciate ogni speranza, voi ch'entrate :-)
(Read Only)pid=20409

No more posts

MatPlus.Net Forum Internet and Computing Natch as Natch can