Hi Steve,
SteveG48 wrote:(6=82)r9c68 - (2=13895)r9c45689 - (5=6)r5c8 -> -6 r24c8 ; stte
Did you notice that it's actually a loop with the same exact eliminations as my Sue de Coc? As an MSLS:
MSNS: {5N8 8N45689 9N68 \ 1389r8 8r9 56c8 2b8} => -6 r24c8, -38 r8c7, -8 r9c79
The difference between this and mine is that this has more cells than digits and the 8 needs two covers, so it's not a DDS (
Distributed Disjoint Subset) like the SdC. That's the significant difference between a DDS and a generic MSNS (Multi-Sector Naked Set), as far as I understand. All DDS are MSNS, but all MSNS are not DDS. What they have in common is Rank 0 and only cells as truths. Or so I think.
Btw, my direct translation into its complementary view doesn't seem exactly efficient:
- Code: Select all
 .-----------.-------------------.----------------------.
| 4  6  37  |  378    9    5    |  1       2      378  |
| 8  9  137 |  137    124  1237 |  3467   *37-6   5    |
| 5  2  137 |  6      14   1378 |  3478    9      3478 |
:-----------+-------------------+----------------------:
| 9  4  2   |  17     5    167  |  3678   *378-6  3678 |
| 7  1  8   |  4      3    69   |  569     56     2    |
| 3  5  6   |  2      8    79   |  79      4      1    |
:-----------+-------------------+----------------------:
| 2  8  5   | *139    6    4    |  37     *137    379  |
| 6  7  4   | *1389  *12  *1238 | *25-38  *1358   389  |
| 1  3  9   |  5      7   *28   | *246-8  *68    *46-8 |
'-----------'-------------------'----------------------'
MSHS: {25R8 246R9 1378C8 1389B8 \ 24n8 7n48 8n45678 9n6789} => the same