The New Sudoku Players' Forum

[ArkieTech]

Code: Select all

 *-----------*
 |.47|..9|5..|
 |..1|72.|.8.|
 |3..|..5|4..|
 |---+---+---|
 |...|...|7.9|
 |.7.|...|.5.|
 |9.8|...|...|
 |---+---+---|
 |..2|4..|..1|
 |.3.|.12|6..|
 |..4|9..|82.|
 *-----------*


Play/Print this puzzle online

[SpAce]

Code: Select all

.-------------------.--------------------.---------------------.
|  28    4      7   | 1368   368    9    | 5    16       2-6   |
| b56   b569    1   | 7      2      4    | 39   8     ad(3[6]) |
|  3     28    b69  | 168    68     5    | 4    1679     27-6  |
:-------------------+--------------------+---------------------:
|  1256  1256   356 | 23568  34568  1368 | 7    346      9     |
|  4     7      36  | 236    9      136  | 12   5        8     |
|  9     1256   8   | 2356   34567  1367 | 12   346    d(346)  |
:-------------------+--------------------+---------------------:
|  5678  5689   2   | 4      35678  3678 | 39   379      1     |
|  578   3     c59  | 58     1      2    | 6   c479    d(4)57  |
|  1567  156    4   | 9      3567   367  | 8    2        357   |
'-------------------'--------------------'---------------------'

(6)r2c9 = (6,9)b1p459 - (94)r8c38 = (436)r862c9 => -6 r13c9; stte

[Ngisa]

Code: Select all

+-------------------+----------------------+-----------------+
| 28     4      7   | 1368    368     9    | 5    16     26  |
| 56     569    1   | 7       2       4    | 39   8      36  |
| 3      28    d69  | 168     68      5    | 4   e169-7  267 |
+-------------------+----------------------+-----------------+
| 1256   1256   356 | 23568   34568   1368 | 7    346    9   |
| 4      7      36  | 1236    9       136  | 12   5      8   |
| 9      1256   8   | 2356    34567   1367 | 12   346    346 |
+-------------------+----------------------+-----------------+
| 5678  b5689   2   | 4       35678   3678 |a39  a379    1   |
| 578    3     c59  | 58      1       2    | 6    479    457 |
| 1567   156    4   | 9       3567    367  | 8    2      357 |
+-------------------+----------------------+-----------------+


(7=39)r7c78 - (9)r7c2 = r8c3 - r3c3 = (9)r3c8 => - 7r3c8; stte

Clement

[SteveG48]

Code: Select all

 *--------------------------------------------------------------------*
 | 28     4      7      | 1368   368    9      | 5      16     26     |
 | 56     569    1      | 7      2      4      | 39     8      36     |
 | 3     a28    a69     |a168   a68     5      | 4     d1679  e67-2   |
 *----------------------+----------------------+----------------------|
 | 1256   1256   356    | 23568  34568  1368   | 7      346    9      |
 | 4      7      36     | 236    9      136    | 12     5      8      |
 | 9      1256   8      | 2356   34567  1367   | 12     346    346    |
 *----------------------+----------------------+----------------------|
 | 5678  c5689   2      | 4      35678  3678   |c39    c379    1      |
 | 578    3     b59     | 58     1      2      | 6      479    457    |
 | 1567   156    4      | 9      3567   367    | 8      2      357    |
 *--------------------------------------------------------------------*


(2=1689)r3c2345 - 9r8c3 = (937)r7c278 - 7r3c8 = 7r3c9 => -2 r3c9 ; stte

[Sudtyro2]

Code: Select all

+-------------------+-------------------+------------------+
| 28     4      7   | 1368  368   9     |  5    16    26   |
| 56     569    1   | 7     2     4     |  39   8     36   |
| 3     c28     69  | 168   68    5     |  4   b1679 c267  |
+-------------------+-------------------+------------------+
| 1256   1256   356 | 23568 34568 1368  |  7    346   9    |
| 4      7      36  | 236   9     136   |  12   5     8    |
| 9      1256   8   | 2356  34567 1367  |  12   346   346  |
+-------------------+-------------------+------------------+
| 5678  d568-9  2   | 4     35678 3678  | a39  a379   1    |
| 578    3      59  | 58    1     2     |  6    479# 457   |
| 1567   156    4   | 9     3567  367   |  8    2     357  |
+-------------------+-------------------+------------------+

(93=7)r7c78 - r3c8 = (72-8)r3c29 = 8r7c2 => - 9r7c2; stte

Or, add a beginning node to get the remote-fin chain for ...
Kraken franken 1-Fish (9)r8\b7 + rfr8c8 => - 9r7c2; stte
9r8c8 - (93=7)r7c78 - r3c8 = (72-8)r3c29 = 8r7c2.

And, as SpAce suggested here one could also write the fish logic as
(9)r8\b7 = 9r8c8 - (93=7)r7c78 - r3c8 = (72-8)r3c29 = 8r7c2 => - 9r7c2; stte

BTW, my understanding of JC Van Hay's comments about "false" fish is now much improved after reviewing my old notes and your recent comments. You are correct in noting that a fish body may, in fact, be false, but JC never seemed to say that it was also no longer a fish at all. Hence, your single-chain logic would still be valid in those cases.

SteveC

[SpAce]

Hi SteveC!

And, as SpAce suggested here one could also write the fish logic as
(9)r8\b7 = 9r8c8 - (93=7)r7c78 - r3c8 = (72-8)r3c29 = 8r7c2 => - 9r7c2; stte

Indeed. However, I recently thought about it some more. As mentioned on Aug 3, I've been wondering if we should use Allan Barker's convention of capitalizing the base sets to make the fish node stand out more within a chain, i.e. (9)R8\b7 here. Even though the '\' is enough to disambiguate it, it's not necessarily obvious at a glance, especially with a mutant fish that looks like normal cell coordinates, e.g. (9)r8\c7. With the capitalization (9)R8\c7 it's much more obviously different. What do you think? Here's a recent example where I used that style (in the hidden solution).

BTW, my understanding of JC Van Hay's comments about "false" fish is now much improved after reviewing my old notes and your recent comments. You are correct in noting that a fish body may, in fact, be false, but JC never seemed to say that it was also no longer a fish at all. Hence, your single-chain logic would still be valid in those cases.

Good to hear that we seem to be in agreement! I also reviewed some of those old fish discussions where JC participated, and I think I learned some more.

The headless fish I mentioned earlier is apparently not the only invalid fish type. It's just the easiest to spot as such because it has fewer cover sets than base sets (or otherwise it has a cover set without any base candidates). That is the case when it's not really a fish at all, and should rather be used as a DP. One reason is that XSudo refuses to see it as a valid construct (tested by StrmCkr), and I think it's the reference set logic implementation we should probably trust (even though I've never used it myself).

However, apparently there can also be fish bodies that look all right, having equal numbers of bases and real covers, but still no solutions. Those can be impossible to spot as invalid without template analysis. Fortunately it doesn't matter. The fish=fin(s) rule works for invalid fish bodies too, so there's no reason to worry about it. (It works for the headless fishes too, but it just makes more sense to see them as DPs.)