| Author |
Message |
Roger J Cooper (Rogercooper)
| | Posted on Sunday, March 30, 2003 - 3:18 pm: | |
I have found a bug in the repetition keyword. It ignores the existence off-board pieces when determining the position is the same. |
Andrew Dabrowski (Dabrowsa)
| | Posted on Sunday, March 30, 2003 - 6:35 pm: | |
See the thread
/discus/messages/3/370.html?1046233401.
In that case I had a game with no off-board pieces,
so that cannot be the only source of the problem. |
Roger J Cooper (Rogercooper)
| | Posted on Sunday, August 24, 2003 - 12:17 pm: | |
I noticed that ZOG2 has still not fixed the repetition bug. Games are still being reported as draws even if the number of off-board pieces is not the same.
In addition, it would be nice to be able to check the number of off-boards pieces. An option in the play menu could be used to display a box with the numbers, types and sides of off-board pieces listed. |
Jeff Mallett (Jeffm)
| | Posted on Sunday, August 24, 2003 - 6:51 pm: | |
This is as designed, not a bug.
Many (most?) games want to start with an infinite number of pieces to drop. Some large initial "off" value is set in the ZRF and decremented after every drop. Even though the "off" value is different in different positions, for practical purposes the game would consider it the same (infinity).
I can see the value of adding a new feature to support both situations -- a new ZRF option for whether or not off-board pieces should be counted as part of the game state. |
Roger J Cooper (Rogercooper)
| | Posted on Saturday, September 06, 2003 - 12:16 pm: | |
I wonder what games you would want to ignore off-board pieces for repetition. The games with unlimited off-board pieces are normally games of position (with no captures) so repetition is impossible. I can't think of a single game with captures and unlimited off-board pieces. |
Jeff Mallett (Jeffm)
| | Posted on Saturday, September 06, 2003 - 4:41 pm: | |
How about Ninuki-Renju? (under the Go-Moku icon in Zillions) It doesn't matter if we can think of existing games though -- if I only tried to support known games then most of the games on this site wouldn't have been possible. |
Jose Carrillo (J_carrillo_vii)
New member
Username: J_carrillo_vii
Post Number: 11
Registered: 5-2008
| | Posted on Saturday, January 17, 2015 - 3:31 pm: | |
To me the bug (or incorrect design feature) is that it considers the pieces off board when it determines Repetition. In my game the dummy pieces always change (because I'm using the dummy pieces to simulate a moves counter), therefore making the Repetition goal useless, as it is designed.
There should be a parameter feature to turn on/off as to whether one wants the off-board pieces to be considered by the Repetition goal, or not. |
Greg Schmidt (Gschmidt2)
New member
Username: Gschmidt2
Post Number: 179
Registered: 1-2007
| | Posted on Saturday, January 17, 2015 - 4:30 pm: | |
Are both the off board pieces and positions specified as "dummy"? |
Jose Carrillo (J_carrillo_vii)
New member
Username: J_carrillo_vii
Post Number: 12
Registered: 5-2008
| | Posted on Saturday, January 17, 2015 - 4:36 pm: | |
I have both.
For the counter, both pieces and positions are "dummy".
But to track the player who captured last, I have a real piece in a dummy position.
Does that make any difference? Thanks. |
Jose Carrillo (J_carrillo_vii)
New member
Username: J_carrillo_vii
Post Number: 13
Registered: 5-2008
| | Posted on Saturday, January 17, 2015 - 4:49 pm: | |
According to the ZoG Rules File Language Reference, as it regards to the Repetition goal:
When comparing positions to see if a repetition has occurred, the entire state of the board is noted, including the current piece attributes. Two pieces of the same color, piece-type, and attributes are considered identical here just as in the game's movement rules, i.e. it does not matter if two identical pieces have swapped places, the position is still considered a repetition. The off-board stores are currently not considered when comparing positions.
- - -
My interpretation of off-board is that the dummy pieces/positions should be ignored.
Is this the right interpretation? |
Greg Schmidt (Gschmidt2)
New member
Username: Gschmidt2
Post Number: 180
Registered: 1-2007
| | Posted on Saturday, January 17, 2015 - 5:01 pm: | |
I don't know if you're tracking piece, which is not a dummy piece, is causing the problem or not. You could try making it a dummy piece and see if that matters or not. Or you could try some experiments with a very simple game with only two positions. I could be wrong, but I believe "The off-board stores" do not refer to off-board positions which you create, I believe it refers to the internal off-board store which Zillions uses to hold pieces that have not yet been "dropped". BTW, I generally have the opposite problem, I want to turn off 3-times repetition as it often gets in the way of games/puzzles for which it's not relevant. |
Jose Carrillo (J_carrillo_vii)
New member
Username: J_carrillo_vii
Post Number: 14
Registered: 5-2008
| | Posted on Saturday, January 17, 2015 - 5:21 pm: | |
Nope. Just tried it. Got rid of all the logic (as a test) to keep track of who captured last, and kept only the counter logic, which uses dummy pieces and dummy positions exclusively, and still had the problem with Repetition not being recognized.
I disabled my counter, and tried again, and Repetition worked.
Fortunately for you, I think that identifying my problem solves yours!
Just keep a changing dummy board, and the repetition will never work in your games/puzzles. |
Jose Carrillo (J_carrillo_vii)
New member
Username: J_carrillo_vii
Post Number: 15
Registered: 5-2008
| | Posted on Saturday, January 17, 2015 - 5:31 pm: | |
I'm not the only one impacted by the bug that potentially solves your problem Greg.
/discus/messages/3/207.html
This issue with the dummy pieces affecting Repetition was already raised way back in 2001. |
Greg Schmidt (Gschmidt2)
New member
Username: Gschmidt2
Post Number: 181
Registered: 1-2007
| | Posted on Sunday, January 18, 2015 - 12:22 am: | |
Yes, I had solved it that way in the past. I have the opposite problem, because for the games/puzzles I'm interested in, repetition detection just gets in the way. I agree that more control over repetition is an important feature. |