Dear retro friends,
I would like to introduce “Jacobi”, a new program that can solve fairy proof games. The home page is at http://wismuth.com/jacobi/ . The program is a web app that runs in your browser and requires no installation.
read more…
It can solve proof games that Popeye struggles with, which by itself is a good enough reason to add Jacobi to your toolbox. There are also new possibilities like allowing undefined pieces in the diagram of a proof game.
That said, it is still far from being a “Natch/Euclide for fairy chess”. Jacobi’s introductory problem is in 9.5 moves, while Natch’s is in 22.5 and Euclide’s in 26.0. But this is the first version and performance might improve over time.
How to try Jacobi for people in a hurry:
– copy any example problem from the home page to your clipboard
– click on “Launch the latest version”
– paste the problem into the box below “Text representation”
– click on “Set problem from text”
– click on “Start”.
I hope that you find it useful! – François Labelle
Great addition to the toolbox of fairy chess composers.
My Patrol chess PG in 14 from 1994 seems to be correct! 🙂
“Julia” is already taken…
There is a programming language with that name… 😉
https://julialang.org/
I would like to draw attention to a feature that I added specially with Julia’s Fairies in mind. It is the output language “py2web format”. Julia probably prefers to receive solutions in this format since JF is using Py2Web to display diagrams and solutions.
If you see something like “{display-departure-file}” in the output, do not remove it! It is a Py2Web annotation that will disambiguate the algebraic notation.
One example of this is https://juliasfairies.com/problems/jf-2016-ii/no-1174/ . Solving the Popeye code from Geoff’s comment in Jacobi gives
1.c7-c8=S 2.a7-a8=Q 3.Qa8-h1 4.Sc8-b6 5.Sb6-d7 6.Sd7-f8 7.g7-g8=R 8.Rg8-g6 9.Kh8-g8 10.h7-h8=B 11.Bh8-g7 12.Qh1-h8 13.Sf8-h7 14.Bg7-f8 15.Qa1-g7 {display-departure-file} 16.Rf5-f6 {display-departure-rank} 17.f4-f5 18.Bc1-h6 19.g4-g5 {==}
which is exactly what you would send to Julia. For problems that don’t work in Py2Web, Julia might prefer to receive “English-S” output. In case of doubt, send both.
Would it be possible to interrupt the solution and continue with it later?
Does the Pause/Resume button do what you want? You have to keep the window open though. One day I might add the ability to resume even after a reboot.
Meanwhile, the 0.2.2 version was launched. It has now Chess960 condition. For example, Tacu’s enigma 17222 (Die Schwalbe 287/2017) is tested in 3 seconds.
Es Jacobi un clon de Popeye?
The first 13 moves of P1109636 (Graefrath) tested by Jacobi in about 18 hours:
Done. Compute time = 64072.550 s
exact solutions (1)
1.h4 c6 2.h5 Qb6 3.h6 Qxf2 4.hxg7 Qxe1 5.gxh8=K Qxd1 6.Kxg8 Qxc1 7.Kxh7 Qxb1 8.Rxb1 Bh6 9.Kxh6 f5 10.Kh5 Kf7 11.Kh4 Ke6 12.Kg3 Kd5 13.Kf2 f4
Maybe somebody should completely verify. The last white move is 14.Ke1.
Nr.325 (PG11.5, Multicaptures, Rãican) is the first PG with this condition checked by Jacobi, version 0.3.1., in about 2 days.