MatPlus.Net

 Website founded by
Milan Velimirović
in 2006

6:03 UTC
ISC 2024
 
  Forum*
 
 
 
 

Username:

Password:

Remember me

 
Forgot your
password?
Click here!
SIGN IN
to create your account if you don't already have one.
CHESS
SOLVING

Tournaments
Rating lists
1-Jan-2024

B P C F





 
 
MatPlus.Net Forum Internet and Computing Helpmate Analyzer – new version and new web address
 
You can only view this page!
(1) Posted by Viktoras Paliulionis [Tuesday, May 7, 2019 13:47]

Helpmate Analyzer – new version and new web address


A new version of Helpmate Analyzer has been released. Helpmate Analyzer now has a responsive/mobile-friendly user interface, and it works on various mobile devices and web browsers.

The new Helpmate Analyzer version has some improvements and new features:
- Animation of a solution.
- Synchronizing of the current position in the diagram and the selected move in the table of arrival/departure effects.
- Highlighting moves that are repeating in different phases.
- Showing all final (mating) positions.
- Showing twins in separate tabs.
- Facilitation of the search in chess problem databases (PDB and YACPDB).
Helpmate Analyzer can identify more than 100 themes.

Helpmate Analyzer is accessed via new web address: http://helpman.komtera.lt/
The old version at http://www.komtera.lt/helpman/ will not be supported and will not be available after some time, so please use the new web address.
 
(Read Only)pid=17297
(2) Posted by Hauke Reddmann [Tuesday, May 7, 2019 22:17]

"Helpmate Analyzer can identify more than 100 themes."
One word: how? :-)

I would be most interested in your methodology, especially
now that I'm studying computer science.
I'm still dreaming that someone does the same job for
twomovers.
Another suggestion for v3.0: Recognizing Antizielelemente.
I tried my only(?) (at least orthodox) h# * and it did
not get the gist of my problem, probably as I am still too
original for AI :-)

Hauke

* SCHWALBE 122, 1990 (Nr.6901), or the even better version:
http://bambam42.de/problem/version.html#S6901
 
 
(Read Only)pid=17298
(3) Posted by Viktoras Paliulionis [Wednesday, May 8, 2019 19:02]; edited by Viktoras Paliulionis [19-05-08]

All themes according to their identification method can be divided into three groups:

1) Themes that do not depend on the content of the play. It is enough to examine the trajectories of pieces, even without knowing the initial position of the problem. Most of these themes are common to all genres (roundtrip, switchback, pendulum, etc.). Even themes like Rehm or Zilahi are purely geometric. You can look at Turevsky's Yacpdb source code to find out how such themes are recognized.

2) Themes that are defined as a set of certain motifs. To identify such themes, you must first determine the motivation of each move and arrival / departure effects (such as guarding, line closing, line opening, self-blocking, pinning, etc.). Examples of such themes are the Maslar, Indian, Leibovici interference and others.

3) Themes requiring not only an initial initial position and a solution to recognize them, but also an analysis of tries and possible alternative ways for solutions. Examples of such themes are tempo play, dual avoidance, anti-critical move, antizielelement. The problem is that the solving programs do not provide tries for helpmates. Finding tries often is more difficult than finding a solution. Another problem is that there are no precise and unambiguous definitions of these themes. These themes are the most difficult for the Analyzer and not all are implemented. Currently, it cannot recognize antizielelement. Can you give a precise definition of an antizielelement?
 
   
(Read Only)pid=17303
(4) Posted by Hauke Reddmann [Wednesday, May 8, 2019 20:24]

Ask the inventor :-)
(My try: Any effect of a white move that must be reversed -
or compensated, watering the definition even more -
by the rest of the solution. Half of the #2 Status Quo group -
do not confuse with the group Status Quo :-) - would fall
under it. White selfpins followed by black unpin, unpinning
black followed by repinning, ...)
 
   
(Read Only)pid=17304
(5) Posted by Viktoras Paliulionis [Thursday, May 9, 2019 09:27]; edited by Viktoras Paliulionis [19-05-09]

I think, that it can be a black move as well. For the purpose of programming such definition of Antizielelemente is too abstract. The definition in the Encyclopedia of Chess Problems is also not concrete.
If we could define all possible cases, then it could be programmed. I am currently seeing the following Antizielelemente cases:
- line closing / line opening,
- unpin / pin (of black piece),
- pin / unpin (of white piece),
- unblock / self-block,
- unguard / guard,
- depart / return (switchback).

The current version already recognizes some of these cases:
- unpin-pin,
- pin-unpin,
- blocking piece replacement.

There are also a number of themes, that by definition also has features Antizielelemente:
- Bukovina theme (unguard by capture/self-block)
- Chernous theme (self-pin/unpin)
- Klasinc theme (switchback)
- Feather 2 theme (unguard/self-block by capture)
 
   
(Read Only)pid=17308
(6) Posted by Hauke Reddmann [Thursday, May 9, 2019 10:12]; edited by Hauke Reddmann [19-05-09]

Yes, I agree that an AZE can be any white loss or black gain
and can be created/destroyed by just one party or both.
Also willingly or forced since AZE is not genre-bound.
(Add flight-take/give to your list. My example given above falls
under this.)
 
   
(Read Only)pid=17309
(7) Posted by Viktoras Paliulionis [Thursday, May 9, 2019 17:31]

I'm not sure if your example (S6901) contains AZE. The helpmate problems, unlike the direct-mate problems, have no weakening and defensive motifs. For example, the first move of the black knight is just an intermediate move into the destination field. It has no effect.
 
 
(Read Only)pid=17312

No more posts


MatPlus.Net Forum Internet and Computing Helpmate Analyzer – new version and new web address