... by: kolumbus

My solution is similar to Karl's (took me some hours):
if H9=4 -> F9=1 -> G7=1 -> no 8 in row G
and H5=7 -> D4=7 -> no 8 in row D -> contradiction to Setti on 8
-> H9=1 -> no 2 in row G -> the rest is basic eliminations

John's "almost UR" in short form: A normal UR 89 on AB69, followed by 3 x-wings: 7 on AB79, 8 on AB69, and 6 on AB78 (after naked pair 78 on EF8).

... by: Klaus

Hi John,
we indeed have different starting positions. i have D2=123 (i see no reason that this must be 1), A4=1235, B2=239.
My argument was not that i tried all combinations of all candidates, but if there were no 5 and no 9 in row A and B, then the 5 values of the 8 cells in AB69 do not occur in any of the related rows and columns (A, B, 6, 7, 8, 9), so that the solution of AB69 cannot be decided in the further course of the puzzle solution, which is a contradiction to uniqueness. but this argument is wrong.
furthermore i see now that 9 can be removed from B2 due to UR AD12 and that 5 occurs in B3 ,so my argument was not only pseudo-clever, but complete nonsens.

... by: John

@Klaus. As you say, there was something wrong with your assumption and I think it is because you did not look at every variation of candidates in the cells you looked at.
I can see the logic that says "if every variation of candidates in some cells gives the 'same result' somewhere else, then that 'same result' must be correct". But I do not think you looked at every variation of candidates in A4, B2 and B4, which is the mistake. For example, A4=1, B4=5 resolves the corner without a UR. Also, I think you have an error in your starting position because I have D2=1 which means B2 cannot be 9.
I have A4=15, B2=367, B4=2359.

... by: Klaus

Thanks John, indeed very interesting corner.
i came with some other argumentation (i think Setti and x-wing) to the same conclusions as you. every cell in AB69 with two candidates, symetrically in AB and all connected to each other. so only two solutions of this 8 cells are possible.
now i was very clever: i argued that only A4=5 or B2=9 or B4=9 could decide the situation in these 8 cells and that otherwise there would be UR-situation. but all of these 3 possiblities lead to the same solution of the 8 cells. so this must be it!
and guess what? its wrong.
None of the 3 possibilities is right and my "UR-situation" is later decided indirectly by a Setti.
so much about being clever.

... by: John

@Klaus. AB69 is an "almost UR" which has 5 candidates for the 4 columns, so isn't a UR. But it contains three URs! For example, AB6 and AB9 are a UR on 89 which means there must be a 7 in AB9, which eliminates the 9 in AB9. The necessary 7s in AB7 and AB9 XWing to remove the 7s in AB8, leaving the 56 naked pair which removes the 6s in EF8 (Note the 2x2 UR on 67, or the 3x2 UR on 678, also makes a 5 mandatory in AB8 which eliminates 5s in EF8). The 6s in AB78 now X-Wing to remove 6s in the rows, and the 7s in AB79 XWing to remove the 7s. The 8s have already X-Winged to remove the 8s in the rows.
It was an interesting corner!

... by: Klaus

very elegant Karl.
i've got something similiar, but not as nice as yours:
E9=2 would force H9=4, and
G7=1, H3=8 and therefore D4=8,D5=7
no number left for H5

@John: whats the UR argument?

... by: Karl

D4 = 8 can be excluded, because

on the one hand, it would lead to D5 = 7,

on the other hand, it would lead to J6 = 2 (Setti's on 8 for row G, Setti's on 2 for column 6) => H9 = 4,

no number left for H5.

... by: E.Bandit

there has to be an error in my candidate list, puzzle has no soltution with my noted candidates, so better forget the posts, i am most sorry myself

... by: E.Bandit

i missed something: setti chains lead to conclusion that
G1-7 is either 1-7 or 3-9.
but 3-9 makes G7=5 does G6=4; G5=7; G2=6 , so nothing left for G1! (567)

... by: uk

I am new on this side; got the settis explained but: what are swordfish and x-wings?

... by: E.Bandit

B3=5 applied to Setti2 = no 2 in G so F7=1
-> 2x2 : CF45 -> eliminates all 2 in B,
Assumption wrong, puzzle solved (i hope so,did not go on yet)

... by: E.Bandit

list of x-wings and swords:
8x2: AB69
6x2: AB78
6x3: FGH123
7x2: AB79

----------------

possibles in EFG7: 5 2
2 or 1
1 5

almost2x2: c45 and G45 (if G7=1)
and F45 (if G7=5)

either way there will be ONE x-wing :)
---------------
next step is the solution(see above comment)

... by: Klaus

Thank you JimKnopf, but in my "stuck-position" this swordfish does not remove anything anymore.

... by: JimKnopf

@Julia:
x-wing on 6 in ab78 eliminates 6 in b2

... by: Julia

After 15 solved cells (logicand Setti) I got stuck Saw no xwing and cant see the swordfish. What is with the 6 candidate in B2?

... by: JimKnopf

@klaus:
fgh123 (6)

... by: Klaus

wow, hard one.
got no way through yet, saw no Swordfish

... by: JimKnopf

lots of settis, tons of x-wings, a swordfish and one chain