The New Sudoku Players' Forum

[ArkieTech]

Code: Select all

 *-----------*
 |...|.1.|...|
 |..8|6.9|7..|
 |.5.|...|.2.|
 |---+---+---|
 |7..|...|..6|
 |.2.|8.5|.4.|
 |..1|...|9..|
 |---+---+---|
 |.1.|.3.|.5.|
 |..9|...|4..|
 |...|.2.|...|
 *-----------*


Play/Print this puzzle online

[SteveG48]

Code: Select all

 *--------------------------------------------------------------------*
 | 3469  c34679  3467   | 2347   1      23478  | 5      6-3    348    |
 | 2    bd34     8      | 6      5      9      | 7     a13    a134    |
 | 1      5      3467   | 347    478    3478   | 368    2      9      |
 *----------------------+----------------------+----------------------|
 | 7     c3489   5      | 134    49     134    | 2     a38     6      |
 |c369    2     c36     | 8      679    5      |b13     4     b137    |
 |d3468  c3468   1      | 2347   467    23467  | 9     a378    5      |
 *----------------------+----------------------+----------------------|
 | 468    1      2      | 9      3      4678   | 68     5      78     |
 |e3568 ce3678   9      | 157    678    1678   | 4     f1367   2      |
 | 34568 c34678  3467   | 1457   2      14678  | 1368   9      1378   |
 *--------------------------------------------------------------------*


3r246c8,r2c9 = r2c2&r6c79 - 3r14689,r5c13 = 3r2c2&r6c1 - r8c12 = r8c8 => -3 r1c8 ; stte

Or, looking at Hodoku, almost finned swordfish, base c379 cover r159 fins r2c9r3c7:

Swordfish = 3r3c3 - r2c2 = r2c89 => -3 r1c8 ; stte

I think SpAce would write this c379\r159 = 3r3c3 - r2c2 = r2c89 except I don't know how to show the fins.

[eleven]

Code: Select all

 *-------------------------------------------------------------------------*
 |  3469    34679   3467   |  2347   1     23478   |  5    da36     348    |
 |  2       34      8      |  6      5     9       |  7     d13     134    |
 |  1       5       3467   |  347    478   3478    |  368    2      9      |
 |-------------------------+-----------------------+-----------------------|
 |  7       3489    5      |  134    49    134     |  2     c38     6      |
 |  369     2       36     |  8      679   5       | #13     4     #137    |
 |  3468    3468    1      |  2347   467   23467   |  9     c378    5      |
 |-------------------------+-----------------------+-----------------------|
 |  468     1       2      |  9      3     4678    |  68     5      78     |
 |  3568    3678    9      |  157    678   1678    |  4     b1367   2      |
 |  34568   34678   3467   |  1457   2     14678   | #1368   9     #1378   |
 *-------------------------------------------------------------------------*

UR 13r59c13: 3r46c8=13r8c8
6r1c8 = (6-13)r8c8 = 3r46c8 - (3=16)r12c8 => 6r1c8, stte

[Sudtyro2]

Code: Select all

              b                         b          b
+-------------------+-----------------+-----------------+
| 3469  34679 3467  | 2347 1   23478  | 5     6-3  348  |c
| 2    b34    8     | 6    5   9      | 7    c13  c134  |
| 1     5    a3467@ | 347  478 3478   | 368#  2    9    |
+-------------------+-----------------+-----------------+
| 7     3489  5     | 134  49  134    | 2     38   6    |
| 369   2     36    | 8    679 5      | 13    4    137  |c
| 3468  3468  1     | 2347 467 23467  | 9     378  5    |
+-------------------+-----------------+-----------------+
| 468   1     2     | 9    3   4678   | 68    5    78   |
| 3568  3678  9     | 157  678 1678   | 4     1367 2    |
| 34568 34678 3467  | 1457 2   14678  | 1368  9    1378 |c
+-------------------+-----------------+-----------------+

Hi Steve,
Here's one possible UFG-style Swordfish convention...
The base(b) and cover(c) sectors are labeled on the grid boundary.

In 3s, a kraken 3x3-Fish c379\r159 + fr3c7 + rfr3c3 => -3r1c8; stte

Code: Select all

Direct fin(#): r3c7                - 3r1c8.
Remote-fin(@): r3c3 - r2c2 = r2c89 - 3r1c8.

The 3x3 notation allows for later Obi-style NxM Fish.

SteveC

[SteveG48]

UR 13r59c13: 3r46c8=13r8c8
6r1c8 = (6-13)r8c8 = 3r46c8 - (3=16)r12c8 => 6r1c8, stte

NIce!

[SteveG48]

Code: Select all

              b                         b          b
+-------------------+-----------------+-----------------+
| 3469  34679 3467  | 2347 1   23478  | 5     6-3  348  |c
| 2    b34    8     | 6    5   9      | 7    c13  c134  |
| 1     5    a3467@ | 347  478 3478   | 368#  2    9    |
+-------------------+-----------------+-----------------+
| 7     3489  5     | 134  49  134    | 2     38   6    |
| 369   2     36    | 8    679 5      | 13    4    137  |c
| 3468  3468  1     | 2347 467 23467  | 9     378  5    |
+-------------------+-----------------+-----------------+
| 468   1     2     | 9    3   4678   | 68    5    78   |
| 3568  3678  9     | 157  678 1678   | 4     1367 2    |
| 34568 34678 3467  | 1457 2   14678  | 1368  9    1378 |c
+-------------------+-----------------+-----------------+

Hi Steve,
Here's one possible UFG-style Swordfish convention...
The base(b) and cover(c) sectors are labeled on the grid boundary.

In 3s, a kraken 3x3-Fish c379\r159 + fr3c7 + rfr3c3 => -3r1c8; stte

Code: Select all

Direct fin(#): r3c7                - 3r1c8.
Remote-fin(@): r3c3 - r2c2 = r2c89 - 3r1c8.

The 3x3 notation allows for later Obi-style NxM Fish.

SteveC

Thanks, Steve. Don't we also need the direct fin r2c9?

Having looked back over some posts, I think SpAce would have written the first node as c379\r159 f:r2c9,r3c7 .

[Sudtyro2]

Thanks, Steve. Don't we also need the direct fin r2c9?

Oops! Yes, a good catch! One can simply add r2c9 to the direct chain's r3c7.
BTW, the Direct Fin(#) chain is not really needed since those fins see the exclusion cell directly.

SteveC

[SpAce]

This was fun. My first coloring proved 26 trap eliminations, yet no cigar. Some combination of them might have solved the puzzle (btte at most even in that case), I'm not sure, but (uncharacteristically) I wasn't interested in trying to fit them into a single move. A contradiction would have solved the puzzle but that wasn't to be found in that coloring (not that I like them anyway). Had to attack from another angle. Found yet another combo trap, but this time it worked.

Code: Select all

.--------------------------.------------------.-----------------------.
|   3469     34679   467-3 | 2347  1    23478 |  5       b36     348  |
|   2      d(3)4     8     | 6     5    9     |  7       c13    c134  |
|   1        5      a467-3 | 347   478  3478  | b368      2      9    |
:--------------------------+------------------+-----------------------:
|   7        3489    5     | 134   49   134   |  2        38     6    |
|   369      2     a[3]6   | 8     679  5     |  13       4      137  |
|   3468     3468    1     | 2347  467  23467 |  9        378    5    |
:--------------------------+------------------+-----------------------:
|   468      1       2     | 9     3    4678  |  68       5      78   |
| d(3)568  d(3)678   9     | 157   678  1678  |  4       c1367   2    |
|   34568    34678   467-3 | 1457  2    14678 |  1368     9      1378 |
'--------------------------'------------------'-----------------------'

(3,6)r53c3 = (6,3)b3p72 - r2c89|r8c8 = (3)r2c2&r8c23 => -3 r139c3; stte

Edit: a cosmetic change.

[SpAce]

Hi Steves! About the fishes...

Or, looking at Hodoku, almost finned swordfish, base c379 cover r159 fins r2c9r3c7:

Finned Swordfish = 3r3c3 - r2c2 = r2c89 => -3 r1c8 ; stte

I think SpAce would write this c379\r159 = 3r3c3 - r2c2 = r2c89 except I don't know how to show the fins.

Ah, spotting that would have been so nice. Anyway, I'm glad to see you show interest in the UFG style! Despite my own preferences, I don't want to sound like a nazi about it In this case I wouldn't actually mind text so much because UFG finned fishes are quite verbose too. However, then it's important to mention that the node is actually a finned fish (colored addition above).

Here's one possible UFG-style Swordfish convention...

In 3s, a kraken 3x3-Fish c379\r159 + fr3c7 + rfr3c3 => -3r1c8; stte

Yes, that's very close to standard UFG. The only deviation I see is the "3x3" which is Obi-fish style. All UFG fishes are NxN (+fins), so I think just "3-fish" or "Swordfish" would be better to avoid redundancy and confusion with Obi-fishes (*). It also displays some things I don't like about the UFG notation. First, it doesn't have a standard for specifying the fish digit, which results in unnecessary verbosity ("In 3s..."). Second, the fin notation is ugly without any separator before the cell coordinates. I would deviate from the UFG on both accounts and write something like:

(3)c379\r159 f:r2c9,r3c7 rf:r3c3

or:

3:c379\r159 f:r2c9,r3c7 rf:r3c3

Any opinions on that? The only difference between the two is how the fish digit is included. The first one is normal Eureka style, but if used in chains, is it too easy to mix with normal nodes? The '\' should help avoid any confusion, though.

(*) [Added: I just noticed this part:

The 3x3 notation allows for later Obi-style NxM Fish.

I see. I still probably wouldn't mix the two, even with that intention in mind. Both are complicated enough in their own right, and have quite different philosophies. I think it's probably best to learn first one (probably UFG) well, and then the other. Obi-fishes are my own preference these days, but the initial context switch wasn't trivial. More about them below.]

Another problem with UFG is the verbose fin notation, which is awkward in chains. I'd write the chain with a 3x4 Obi-fish instead, since adding the b3-cover takes care of both direct fins (and doubly covers the victim, which neutralizes the extra cover sector). It simplifies (or at least shortens) the fish node considerably:

(3)c379\r159b3 = r3c3 - r2c2 = (3)r2c89 => -3 r1c8

Also, since the chain uses only fish digits, we could write the whole thing as a complex fish without any chain. For that there are several possibilities (check Hodoku), but the one below one was my direct translation from the chain.

Exo+Endo-Finned Mutant Jellyfish (UFG):

(3)r2c379\r159b1 f:r2c8,r3c7 ef:r2c9 => -3 r1c8

I just added the strongly linked r2 as a base, and the weakly linked b1 as a cover. That makes r2c9 an endofin because both r2 and c9 are bases, but that's no problem because it sees the victim too.

The same can be written more shortly as a 4x6 Obi-Fish (my preference again):

(3)r2c379\r159b133 => -3 r1c8

Note the double-cover b3 due to the endo-fin (which needs two covers). This could be avoided by seeing r2c9 as a set triplet and following Allan Barker's rules about them, but that's more complicated than just adding the cover twice (which works because it covers the victim too causing it to be triply covered, overpowering the two extra cover sectors).

Btw, Obi-Wahn's fish-arithmetic (magic) can be used to simplify that to get this:

(3)c379b2\r1359b3 => -3 r1c8

...which translates into a considerably simpler UFG fish too:

Finned Franken Jellyfish: (3)c379b2\r1359 f:r2c9 => -3 r1c8

fish-arithmetic steps: Show

Code: Select all

r2c379   \ r159b133    : initial fish
r2c379b2 \ r159b1233   : b2 added to both sides
r2c379b2 \ r112359b3   : b123 <-> r123
c379b2   \ r11359b3    : r2 removed from both sides
c379b2   \ r1359b3     : redundant r1 removed (no endofins -> no need)

Having looked back over some posts, I think SpAce would have written the first node as c379\r159 f:r2c9,r3c7 .

Yes. Like I said, I like to use a colon with those 'f', 'ef', and 'rf' prefixes for readability. SteveC's style is more in line with standard UFG, though, and Hodoku uses that too (without the '\', though). I just think that something like 'efr2c9' looks weird without any separator. (Then again, I mostly use the Obi-style which avoids that problem altogether.)

Edit: added the simpler fish obtained with Obi-Wahn's fish arithmetic.

[rjamil]

Hi SpAce,

(3)c379\r159 f:r2c9,r3c7 rf:r3c3

or:

3:c379\r159 f:r2c9,r3c7 rf:r3c3

Any opinions on that? The only difference between the two is how the fish digit is included.

Why not using @ with fish digit?

R. Jamil

[SpAce]

Hi rjamil,

Why not using @ with fish digit?

Because I already use that for other purposes (to mark embedded end-points in chains or nets), though it's not standard (or necessarily good for that). I don't think there's any standard use for that symbol, in any of our notations, but I don't quite see how it would help here. It's not even in line with your own use of it, so it would create yet another different meaning. As you know, I'm not a big fan of overloaded meanings for terms or symbols.

My first suggestion is the most in line with Eureka, but I'm not sure if it creates more or less confusion that way because Eureka doesn't really cover fish notations (hence the other option -- and of course any other suggestions are welcome as well).

[rjamil]

Hi SpAce,

As you know, I'm not a big fan of overloaded meanings for terms or symbols.

I also like simplicity but not overloaded complexity for shortening the terms and/or symbols.

R. Jamil

[SpAce]

As you know, I'm not a big fan of overloaded meanings for terms or symbols.

I also like simplicity but not overloaded complexity for shortening the terms and/or symbols.

That's an overloaded meaning for "overloaded"

[SpAce]

UR 13r59c13: 3r46c8=13r8c8
6r1c8 = (6-13)r8c8 = 3r46c8 - (3=16)r12c8 => 6r1c8, stte

NIce!

Indeed! A very nice find.

Btw, once again the placement conclusion implies it could be simplified. How about just:

6r1c8 = (6-1|3)r8c8 == 3r46c8 => -3 r1c8; stte

That's probably the shortest non-fish solution available.