Längstes Hilfsmatt in illegaler Stellung ohne Berücksichtigung von Dualen
Vorbilder waren die P0569739, P0569740, und die P0569741.
Weiterentwicklung der P1372242
Viktoras Paliulionis: Duals in helpmates are not allowed, so the problem must be marked as cooked. (2020-08-24)
A.Buchanan: How bad are the cooks? (2020-08-24)
Olaf Jenkner: 21 bishops in helpmates are not allowed, so what? (2020-08-24)
Viktoras Paliulionis: The flag "cooked" and the keyword "Illegal position" are useful in queries, e.g. for searching only correct problems. The database must contain accurate information. I think it would be more accurate if this problem were assigned to another genre, for example, to "Fairies". Then the flag "cooked" could be removed. (2020-08-25)
A.Buchanan: Hi Viktoras, thanks for your thoughtful feedback.
The situation here is similar to that of proof games, where there are both non-unique & unique categories. The authors do not assert that this helpmate is unique. But the proper category does not exist (and probably never will). So some inaccuracy of classification is unavoidable here. The question is just which inaccuracy do we prefer. My own feeling is that it's a greater distortion (and harsher to the composers) to mark this as cooked than to live with it as a flawed h#.
The fact that the position is illegal is an independent issue, to me.
The genre "Fairies" should not be a dumping ground for problems that we can't classify. It's basically for non-orthodox rules, boards, pieces etc and for some non-orthodox stipulations e.g. curiously h=. But I don't feel that the number of solutions can push a problem from orthodox into fairy: again I feel that's a separate dimension.
This is a difficult classification example, and people's opinion may well differ! :-) (2020-08-25)
Henrik Juel: This is an extremely special situation, and I generally agree with you, Andrew
'C+ cooked' is a silly label combination, and I would suggest no such labels at all
This would make the authors marginally happier, and it would not hurt an unsuspecting user looking for correct problems, as long as she includes 'cplus' in her search string (2020-08-25)
Viktoras Paliulionis: It's actually strange to see 'C+' when the solution is not unique. I suggest changing the stipulation of the problem to "non-unique-h#" (by analogy with "exact-h#"), then user will not be misled. (2020-08-25)
A.Buchanan: The stipulation belongs to the composers: it is part of their artistic creation. In most cases it's trivial, but otherwise it was selected by them with intent.
What we should have in PDB is a Unicode stipulation which is sacrosanct but is not used to drive the animation engine. If the stipulation includes text, there should be translation (DE+EN+FR), under moderatorial control, driven from canonical vocabulary, and used to inform the animation. This would allow the moderators to impose their own valid judgements, whimsical prejudices etc, which can indeed be useful. And it allows anyone to see the gap between the actual stipulation and the PDB interpretation.
Unfortunately, we don't have this situation, and so moderators can if necessary over-write the authors' actual stipulation! What brazen cheek! :-) On occasions when I have done this in a non-trivial way, I try to retain the Originalforderung in the comment field (not a usercomment field), and to consult the composers if they are still alive.
Particularly, the composers should not be forced to restipulate in order to compensate for some perceived defect in the database classification system. In the case of this problem, I've asked what the duals actually are, but no-one seems to know.
Probably best if I quote the Codex:
Article 9 – Cook
A chess composition is called cooked if it has a solution that differs in its first move from the author’s solution.
Article 10 – Dual
A dual is said to occur if, after the first move, there is more than one method of satisfying the stipulation.
Article 13 – Unsound Chess Compositions
(1) Subject to paragraphs (2) and (3), a composition is unsound if it is cooked or has a short solution or no solution.
(2) Help-play compositions are unsound also if they are dualized, except that in the final move a promotion into different pieces having partially the same power (for example queen/rook or queen/bishop) may be tolerated.
(3) Studies...
Bottom line: so yes a dualized helpmate is unsound, but we have no way to record that here: "cooked" is a much narrower kind of defect. And to my mind it's also a very strong word: I also object to its use in PDB in cases where we have not established that the diagram is correctly reproduced.
The composers should not suffer for the defects of the database! If the dual is severe, then possibly "cooked" is least severe distortion. If the defect is minor, though, then the term "cooked" is a slur :-) (2020-08-25)
James Malcom: Viktoras, when this was composed, everyone involved knew what they were doing. You can reqd the entire three page thread with the original problem here-https://www.matplus.net/start.php?px=1598358654&app=forum&act=posts&fid=gen&tid=2423 (2020-08-25)
James Malcom: But I agree that lableing at as “non-exact” is the best possbilr compromise here. I’ll PM Arno to see what he thinks. Andrew, there actually is precedent for the doing this. The longes known legal helpmate with duals. a #48 by Fabel-see P0569740, has the stipulation h#48 (nicht exakt!), which is what would work here. (2020-08-25)
Viktoras Paliulionis: I think the term "non-unique" is more accurate than "non-exact" for this situation. The term "exact-h#" is a legal term in the PDB. It means the exact number of moves (neither less nor more), but not duals. Therefore, "not-exact" means the usual h# stipulation. "Not-unique-h#111" would mean "any solution in 111 moves". (2020-08-25)
A.Buchanan: Totally cool if composers want to rename :-) I agree with Viktoras that "non-unique" is what we are looking at here, with the current sense of "exact". Maybe should revisit Fabel's stipulation too (while retaining his Originalforderung ofc). In the stipulation we don't need to stick them together. "Non-unique h#111". For now we could have a single keyword "non-unique h#'. (2020-08-25)
James Malcom: I went ahead and updated the stipulation for P1372242. (2020-08-25)
A.Buchanan: ...and maybe I've changed my mind. I can live with it as cooked until the right category appears :-) (2020-08-25, hidden) more ...
comment