The New Sudoku Players' Forum

[9X9]

Just a brief note to say thanks for the interesting time that I've had as a member of the forums and that I won't now be posting here for the foreseeable future. A house move (T minus 2 weeks and counting), a new baby in the new house (T minus 5 weeks and counting), pictures to be painted for an exhibition and a book not being written as fast as I would like, add up to too many other demands on my time.

At least, with 120 posts and my 100th Pappocom Very Hard just solved, I think I've won my sudoku spurs.

If I may just leave you with a final Forums thought. When obvious "newbies" pose queries, by and large a gentle push towards our great friend Angus is all that is required. Well-intentioned personal "advice" and expansion of definitions, especially when one helper's words conflict with another's, often seems to me to be destined to cause more confusion than existed in the first place. Let AJ do the talking.

[lunababy_moonchild]

Just a brief note to say thanks for the interesting time that I've had as a member of the forums and that I won't now be posting here for the foreseeable future. A house move (T minus 2 weeks and counting), a new baby in the new house (T minus 5 weeks and counting), pictures to be painted for an exhibition and a book not being written as fast as I would like, add up to too many other demands on my time.

At least, with 120 posts and my 100th Pappocom Very Hard just solved, I think I've won my sudoku spurs.

If I may just leave you with a final Forums thought. When obvious "newbies" pose queries, by and large a gentle push towards our great friend Angus is all that is required. Well-intentioned personal "advice" and expansion of definitions, especially when one helper's words conflict with another's, often seems to me to be destined to cause more confusion than existed in the first place. Let AJ do the talking.

Congratulations on the house move, new baby and spurs.

I'll miss ya

Thanks for letting us know

Luna

[emm]

Yes, thanks for signing out, 9x9 - I would've wondered!

It's been a bit up and down here at times, but it certainly has been interesting. I take your point about posts – I’ll think of it whenever I’m getting trigger happy with the submit button.

Good luck with everything, especially little 0x0!

[MCC]

I know you've got a lot of stuff on your mind at the moment but, don't forget to feed the fish, and if you come across a plastic fish cruelly stuffed inside a christmas cracker, let it go free.

Good luck with everything.
MCC

[PaulIQ164]

I hereby echo the above comments. Goodbye and good luck!

[CathyW]

Me too! Hope all goes well and that you do return in due course.

[9X9]

I couldn't possibly pull the last branches across the mouth of the cave and hunker down for some serious "hibernating", without first of all thanking you all for the kind words above. It's very touching and a demonstration, if one were needed, of the true fellowship of the grid.

Also, because I might otherwise be doing that very fellowship a disservice if I wait until after the "big sleep" and specifically because I know how initially hard it can be for newcomers to grasp the more complex of the solving methods, I'll risk touching on a few technical points which have been prompting themselves in my mind of late but which I would normally have subjected to longer personal revue before voicing or rejecting them.

1. Our principal referral point for basic solving technique queries is www.angusj.com and the explanations there are very good. However, a suggestion for improvement that I would make is the addition, perhaps as "Locked Candidates 3:", of a definition and example of what our other main referral point, the equally praiseworthy www.simes.clara.co.uk, calls Block / Block Interactions. I personally call these "X-Locks" by virtue of their X-Wing "feel".

2. With the boot on the other foot, I think simes makes no mention of "locked candidates" ('LC') as such, although his "Block and Column / Row Interactions" does cover part of the LC ground covered by angus. Perhaps we could have consistency of content and terminology between the two sites?

3. In the second of the examples given by simes in his detailed explanation of Block / Block Interactions, he says:-

"Secondly, in the example below, the cells marked with * are the only cells in blocks four and six that can contain a 2. This means that 2 can be eliminated from the candidates for the marked cells in block 5 (sic)."

However, block five shows no marked cells.

Perhaps a better way to correct this than by now marking the appropriate block five cells in some way, would be to leave the example as it is but to alter the narrative to read:-

"Secondly, in the example below, the cells in rows four and six marked with * are the only cells in blocks four and six that can contain a 2. This means that 2 can be eliminated from the candidates for rows four and six in block five."

I'll point angus and simes at this post via a PM and then I'm off back to the cave.

[simes]

However, block five shows no marked cells.

They're marked by being in a different colour. Highlighted if you will. (But I've since realised that the colour used isn't too obvious on a LCD screen such as on a laptop.) I've changed the wording as per your suggestion. Thanks

S

[Pat]

re what simes calls Block / Block Interactions:
let's look at his examples -

1. the 3 for c6 is in box 8;
   eliminate 3 elsewhere in this box.
   i.e. Column-Box interaction.
   
2. the 2 for r5 is in box 5;
   eliminate 2 elsewhere in this box.
   i.e. Row-Box interaction.

- Pat

[angusj]

1. Our principal referral point for basic solving technique queries is www.angusj.com and the explanations there are very good. However, a suggestion for improvement that I would make is the addition, perhaps as "Locked Candidates 3:", of a definition and example of what our other main referral point, the equally praiseworthy www.simes.clara.co.uk, calls Block / Block Interactions. I personally call these "X-Locks" by virtue of their X-Wing "feel".

I believe that the Block / Block Interactions technique as discussed by simes is superfluous as it is simply the reciprocal of what he calls Block / Row and Block /Column Interactions. Try out the examples simes has on his 'Block/Block' webpage in Simple Sudoku and hopefully you'll understand better.

[Pat]

re what simes calls Block / Block Interactions:
let's look at his examples -

1. the 3 for c6 is in box 8;
   eliminate 3 elsewhere in this box.
   i.e. Column-Box interaction.
   
2. the 2 for r5 is in box 5;
   eliminate 2 elsewhere in this box.
   i.e. Row-Box interaction.

here are the updated links for simes ( SadMan Software ) --

Solving Techniques

Block / Block Interactions

so i thought i understood his terminology --
but !

Well you say the puzzle can be solved with block/block interaction, but the interactions you refer to are known to me as block/column/row interactions. Which is why the puzzle isn't solved with this technique disabled.

My distinction is simply that if the candidates in one of the blocks are entirely within a single row or column (wrt the other block), then it's BCR, if the candidates are in two columns or rows in both blocks then it's BB. It's only the candidates that are considered, not any big numbers. It's a bit arbitrary, but it works for me.

What I meant by "moot point" was that these two techniques are closely related, and one man's BB is another man's BCR.

[daj95376]

Pat,

I've been following your exchange with Simes in the Programmers Forum. I've also read his description of a Block/Block Interaction. Your example and arguments do not match his description IMHO. However, you have a nice example of a Box having concurrent Column and Row Interactions for the same value. This was Simes argument, I believe.

AngusJ,

Unfortunately, Simes webpage examples are only puzzles and there's no indication of where he thinks a Block/Block Interaction occurs. However, on the first example, where Simple Sudoku finds a Locked Candidate 2, there's an X-Wing for the same value ... and an X-Wing restricted to two boxes qualifies as a Block/Block Interaction as I understand Simes description. Furthermore, the X-Wing produces the same eliminations as your Locked Candidate 2.

[tarek]

there was a brief discussion about this same topic about a year ago on this forum which concluded that block block interaction was basically the same as line block interaction........

it is how you visualize the candidates in the block that determines how you see them or name them......

bottom line.... if you use box line interactions .. no need to use block block interactions.

tarek

[Pat]

interactions between a box and a line
occur in 2 different ways --

1. box-to-line exclusion ( as a fish: b\r or b\c );
   some call it Locked Candidates 1 or Pointing.
   
2. line-to-box exclusion ( as a fish: r\b or r\c );
   some call it Locked Candidates 2 or Claiming.
   
   
these 2 often co-exist,
giving me the choice of using whichever i prefer.

however, they are different methods,
and in many puzzles they do not co-exist.


well, i thought i understood the terminology used by simes --
• Block and Column / Row Interactions
• Block / Block Interactions
-- somehow i matched his 2 terms with the above 2,
and it seemed to fit. consider #9 --

Code: Select all

..6....8.43..1..7....2....5....9.1...2.8.7.4...7.3....2....1....9..8..34.6....5..



-- it can be solved by line-to-box ( and cannot be solved by box-to-line );
and this is matched by simes ( SadMan Software ):
enable only "Block / Block Interactions" (and "singles") and it solved the puzzle,
enable only "Block and Column / Row Interactions" (and "singles") and the puzzle is not solved.

this translation seemed to fit -- until recently.




if you use box line interactions .. no need to use block block interactions





not so! see how SadMan Software behaves in the above example.

[ronk]

Code: Select all

..6....8.43..1..7....2....5....9.1...2.8.7.4...7.3....2....1....9..8..34.6....5..


if you use box line interactions .. no need to use block block interactions

not so! see how SadMan Software behaves in the above example.

I agree with tarek. It appears Sadman Software v2.9.1 has a BUG in its line\box interaction code. It does not detect the r4\b5 1-fish (*) -- the 6s in row 4 being only in box 5.

Code: Select all

 5    1    6    | 34   7    9    | 34   8    2
 4    3    2    | 5    1    8    | 69   7    69
 89   7    89   | 2    46   346  | 34   1    5
----------------+----------------+---------------
 3    8    45   |*46   9   *2456 | 1    25   7
#169  2    159  | 8    5-6  7    |#69   4    3
#69   45   7    | 1    3    245-6| 8    25  #69
----------------+----------------+---------------
 2    45   3    | 9    45   1    | 7    6    8
 17   9    15   | 67   8    56   | 2    3    4
 78   6    48   | 347  2    34   | 5    9    1

It's block/block interaction code, however, does detect the b46\r56 x-wing (#).