dailysudoku.com

[alanr555]

As a resident in UK, I have been used to published puzzles being able
to be solved using "standard" techniques but recently I have had an
extended stay in Spain and am asking whether the same standards are
applicable worldwide.

I purchased four editions (29-32) of a "Sudokus" book (for one Euro
each!) published by Ediciones Saldana SA - DL-M-48019-2008. Each
contains over 160 puzzles accorded easy,medium,difficult gradings.

The easy puzzles seem easy to solve and I have been able to solve
most of the "difficult" puzzles using expanded mandatory pairs. The
medium-grade puzzles are quite a nightmare however.

I thought that I was losing my touch as several had to be abandoned
as "unresolved". Only four of nineteen came out successfully. Thus, I
decided to submit them to the Daily Sudoku on-line process. I found
that the unresolved puzzles were graded as "Too Hard" but that the
D/S methods solved the same cells as I had by hand.

Items found included non-unique rectangles and a mutual dependency
of two digits in six cells plus plain duplicate solutions. Typical was the
puzzle 81 in book 32. I was faced with a choice of 1 and 5 in row 1
columns 1,2. The book gave 1,5 as the solution but 5,1 also works!
004300068; 962001400; 000000100; 200057000; 090604205;
000000010; 070000006; 601090027; 003060081.

Are there any international standards to which publishers generally
adhere? Are they available also in Spanish translation? Am I justified
in being shocked at "suspect" puzzles being published in this way?

I accept that these books were "remainder" stock in an "everything
one Euro" store but does that justify the publication?

We have just received a post from Nataraj on the generation of Sudoku
puzzles. This confirms that many invalid puzzle grids can be generated
by a random process BUT that discernment is applied to select those
worthy of publication (different standards for different destinations). It
would seem that the Spanish publishers were content just to know that
"a" solution exists - irrespective of whether it is unique or capable of
emerging from diligent inspection without recourse to trial and error.
Would this be allowed in an English-speaking environment? Would it
be possible to allege that the publication (if in UK) was unfit for purpose
in terms of the 1979 Act of Parliament on that matter?

Actually, I do not regret my four Euros as I have come to appreciate all
the more the care that goes into my usual sources of Sudoku material
- but I would like to inform Saldana of the error of their ways if, indeed,
international standards do exist.

Have others encountered "defective" puzzles and, if so, what has been
done about them?

In Friendship
Alan Rayner
(now on the Atlantic island of El Hierro)

[enxio27]

Alan, I don't know whether or not you are still around, but I'm just now catching up on the forum after a long absence (simply became busy with other things).

No, there is no world-wide publishing standard, and not even one in the English-speaking world. Unfortunately there are many books, publications, Web sites, and software programs "out there" that produce/publish puzzles that are invalid (multiple solutions). In fact, there is a thread on another forum to highlight such publishers/puzzles in order to warn others away from them:

http://forum.enjoysudoku.com/sudokus-of-shame-t3071-75.html

As I stated on that thread, far too many people who fancied themselves programmers jumped on the sudoku bandwagon when it got popular (around 2005), even though they themselves had no clue about what a sudoku puzzle should look like or how they are constructed. As a result, there are/were many sites and apps out there that generate sudoku-like objects that don't qualify as real sudoku. Bottom line: Buyer beware.