Die Schwalbe

1 problem(s) found in 5992 milliseconds (displaying 1 problem(s)). [PROBID='P1348377'] [download as LaTeX]

1 - P1348377
Ladislav Packa
Vychodoslovenske noviny 1983
P1348377
(6+9) C+
h#2
2 Lösungen
1. axb3ep fxe3 2. Td5 Sb2# (falls zuletzt b2-b4)
1. exd3ep Kxa4 2. Td4 Sxe3# (falls zuletzt d2-d4)
play all play one stop play next play all
sollte die Aufgabe das KW "RV" bekommen?
Henrik Juel: C+ by Popeye 4.61 (after analysis)
Last move was b2-b4 or d2-d4, but we cannot tell which
RV would not hurt, I guess
Some people consider these as one solution, with variations depending on the past (2018-03-23)
A.Buchanan: I am mostly a fan of clarity in the conventions, even at the cost of breaking a problem or two, because problems can fail if people have differing or vague understandings. Werner Keym's modification to the Codex in 2009 was a big step forward. His magisterial paper explaining how the conventions all work after 2009 is a must-read at the Retro Corner site. However, the wording of his added convention is like a speech by Theresa May, and a mess was made of the retro glossary which to this day has not been cleared up. The biggest hole in the Codex is that it doesn't say how to treat a problem which is impacted by a change in the Codex. My guiding principle is that we should always treat such a problem kindly so that it makes the most sense. Anyway, here is what the Codex would imply:
(1) Under RS, which was the default convention in 1983, there is no solution even if the stipulations says that there are two solutions, because you can never prove that a particular double hop was the last move.
(2) Under PRA, which is now the default, there is one solution which has two parts. Everything works perfectly: just remove the "two solutions" from the stip.
(3) Sometimes RV is used as just a synonym for PRA, but it had another meaning. Keym retired the term altogether and, where the distinct meaning is intended, replaced it with SPRA: this operates just like PRA, but en passants are reversed. The default is that they *can* be played, just like castling. But here, if SPRA was specified, it would work just like PRA does: one solution with two parts.
I would rewrite Werner's convention as: "PRA is the default; if that doesn't give a solution, use RS. SPRA or AP would need to be stated explicitly. RV is deprecated."
So here I believe the original problem should have had PRA (or RV) as the stipulation, but now it needs nothing. I think no damage is done by updating to the 2009 version: this makes the problem as accessible as possible in the distant future. I really like this problem: and I think that PRA is a lighter and more natural setting for these 2 model mates than the later Than version.
There is so much confusion over these things that I thought it's worth going into some detail here: I hope it was of interest.
I would be grateful to hear the author's intention and wishes here. (2018-03-23)
Ladislav Packa: Andrew: At the time of publication it was easy for me. Our (Slovak) terminology referred to similar cases as "alternative solutions". I do not know why this option did not appeal to the column editor. He himself added to the problem stipulation 2 solutions. And if I have to express my opinion sincerely, I do not really care about the correct designation. Importantly, the author and especially the readers like it. (2018-03-23)
Mario Richter: @Ladislav+Andrew: Just to make that clear: I didn't want to change the stipulation, for historical correctness it should be as it is. My question was about adding a keyword that describes some problematical content realized in the problem, so that future users of the PDB interested in exactly that topic have a chance to find it by keyword search ... (2018-03-23)
A.Buchanan: @Ladislav: you're the boss! Your historical stipulation will remain to baffle & entertain the robot & alien solvers thousands of years from now when humanity is long gone and new races battle over a tired Earth. Would you object if this problem acquires the PDB keyword "PRA"? (2018-03-23)
Ladislav Packa: In the interest of rescuing the tired Earth I will probably be forced to agree :-)) (2018-03-23)
Henrik Juel: I was pleased to see a happy ending to this debate
But I do not like the fact that Codex revisions make lots of older problems 'incorrect', although they have been enjoyed by solvers, who care little about the Codex
Also, I do not like that 'A posteriori' must be stated explicitly; this gives away the show, as the solver then knows that the key is an ep capture, and a castling must follow to legitimize it (2018-03-23)
A.Buchanan: @Henrik: great to understand your point of view. In 1983 this problem would have been 'incorrect', because RS was the default. W.Keym's revision made it nearer to being correct: the only nit is whether this is 2 solutions or 1 solution with 2 parts. Not sure how I feel about that nit myself.
I honestly don't know any problem which W.Keym made incorrect. It would have to be an RS problem which also has a PRA interpretation. The genius of his terribly-worded revision is that it doesn't affect casual solvers, but it makes things more rigorous for serious solvers. Do you know any such problems? (2018-03-24)
comment
Keywords: En passant as key (2), Model mate (2), Partial Retro Analysis (PRA)
Genre: h#, Retro
Computer test: Popeye 4.61 (after analysis)
FEN: 3r4/1pp5/8/KP6/pPkPp3/n1r1p3/5P2/3N4
Input: Mario Richter, 2018-03-23
Last update: Olaf Jenkner, 2018-03-23 more...
Show statistic for complete result. Show search result faster by using ids.

https://pdb.dieschwalbe.de/search.jsp?expression=PROBID%3D%27P1348377%27

The problems of this query have been registered by the following contributors:

Mario Richter (1)