Cenoman from France

Good to see you here.

Just an explanation on my name : Cenomani were a Gallic people living originally in the area were I was born, not far from the French city 'Le Mans' According to Caesar, 5 000 of them were fighting against him in Alesia (52 before JC)
I ignore totally if I am one of their descendants...

Hi cenoman. Here is an example of a puzzle where ef7=17,ef5=12, and ef4=57 and yet there is a unique solution. It was created by Neil from UK-Hertfordshire, whom I have befriended through this web site.He rarely posts nowadays.

900306007000200080280500904567009002000000070000800463400000090003040005850100000

Be st Regards, Alfred.

Cenoman - that's a very interesting explanation of your name.
My grandfather came from France but I'm not sure now what area.

Cenoman - thank you for letting us know that the 'tough' was back up and working.

Thank you Cenoman, I went there twice with the old family friends and absolutely loved it.
A Sudoku member, GannieMo, who lives in France, visited us there and I went back to stay with her and Steve for a couple of days near Bordeaux. I love the old medieval villages.

Hello Cenoman,

I always follow your elegant solutions in order to learn from them.
In my solution to today (Sept 26, 2020) tough puzzle I had a uniqueness argument as follows:

.---------------.----------------.------------------.
| 14 8 7 | 36 2 369 | 14c 39 5 |
| 25 6 349 | 1 57 379 | 8 3479b 29 |
| 125 135 39 | 4 578 3789 | 27 1379d 6 |
:---------------+----------------+------------------:
| 6 2 15 | 578 4 178 | 3 159 -189 |
| 7 135 34 | 258 9 128 |*156 *1456a -18 |
| 145 9 8 | 35 6 13 | 145 2 7 |
:---------------+----------------+------------------:
| 9 17 12 | 278 78 4 |*56 *56 3 |
| 8 4 6 | 9 3 5 | 27 17e 12f |
| 3 57 25 | 267 1 267 | 9 8 4 |
'---------------'----------------'------------------'

Unique Rectangle gh3, gh5 (56) =>-1 i56

( we must have h5=4 == [g5=1 == h5=1]
if h5=4 => h8<>4 => g9=4 => h6=1 => h2<>1 => i2=1 => -1 i56;
if h5<>4 => [g5=1=h5=1]=> -1 i56 )

Would you be so kind as to let me know how could I write it in a better way. I am asking this because I know of your expertise on the subject!
Thank you in advance.
Regards,
jco

Hello, Cenoman,
Thank you for both information (the limit and the delete feature).
I think in your solution today one part is missing (in the typing)
i146 before loop, right? As usual, it is a neat move!
A great week for you!
jco

Hello, Cenoman
Hodoku Tutorial on Sudoku Techniques was my first introduction to Sudoku. I used the program a lot for training (to find basic patterns like hidden/naked triples, skyscrapers, empty rectangle, fishes, etc). Hodoku's learning mode is not so good for training more advanced patterns like ALS patterns. When I noticed this limitation a while ago, I started to study the moves made by experienced players. A move like the one you found today (Sept 28), Hodoku is unable to find. I always like it when this happens! I mean, I like when a very creative move is still beyond Hodoku capabilities. Hodoku also has very old buggy notation. I use Hodoku to check eliminations after I solve a puzzle because a couple times I miss something. Anyway, I like the program a lot because it helped me to start learning. I also like Sukaku Explainer because it solves puzzles that Hodoku can't. I recall one of the mith's potential hardest could not be solved by Hodoku (and of course, it was solved by you and other experienced players, e a few moves!). Sukaku Explainer solves it in a very long series of forcing chains. By the way, both programs run in any operational system (because are written in Java).
Of course, moves like the one you have found today are well beyond my reach, but I do understand the move after analyzing it and I understand (and makes sense) that need a lot more experience. Fortunately, now I am able to improve little by little (and have fun doing it) by studying moves made by masters like you! : )
Regards,
jco

Hello, Cenoman
I liked the first move (Kraken) in your solution to the very tough puzzle of Oct 1.
Regarding the first chain in the Kraken, for the part
(1)f2 = [(4)d6=f4-(4=2)f2-f78=(2-8)d8=(8)d6]
my understanding is that to justify
f2<>1 => [(4)d6 == (8)d6 (due to mini-chain)]:
{ f2<>1 implies (1)e2==(1)g2 and
e2=1=> g2=9 => df2=24 (locked) => [d1=57, d3=9]=> d159=567 (locked) => d6=48, while
g2=1 => e2=9 => df2=24 (locked) => d135=567 (locked) => d6=48 }
Did I miss a shorter/better way to justify that part? I am under the impression that I am missing something simple.
Regards,
jco

Hello Cenoman

I have studied your comments. As always, very complete, clear and precise. I also could not explain my blunder : )
That Kraken was very clearly presented and I should have understood it without making mistakes. I have promised to myself it will be my last mistake (lets see : |) at least I'll try to triple check always recalling that elementary mistake).
Your explanation was very positive motivation for me to start a comprehensive study on
Embedded AICs,
Kraken,
TMs (have difficulty building one from scratch for complex moves).
I also like a lot when a chain has in it things like fishes, URs and other well-known patterns.
Anyway, an excellent week for you, and thank you!
Regards,
jco

Hello, Cenoman

I read your masterful explanation on my questions and everything was crystal clear after I finished reading. Your explanation also helped to see the beauty of the subject! : )
I will be studying it in detail to get a firm grasp on the concepts and ideas you explained.
It is very nice that there are so many examples from previous puzzles!
(later, I can always practice using them)
Thank you very much!
A great weekend for you!
jco

Hello Cenoman

Beautiful Kraken first move for a difficult puzzle (posted Oct 10)!
I noticed, while studying it, two identical tiny typos.
It has to do with r345 instead of r234.
Anyway, a powerful first move! I have no idea how to find such a move!
Good day!
jco

Hello Cenoman,

I have studied the first part on that long chain with AUR in it.
All very clear! I enjoyed specially the proposed alternative using other guardians, and the final proposed chain is really nice! : )
I will study carefully now the second part (complicated chain with two | in it).
Thank you very much!
Best regards,
jco

Hi Cenoman. I have posted another comment on your amended solution for Oct. 7. I think if the internal c9=1
does not help with the solution, then no external cell can help either.
Best regards,
Alfred.

Hi Alfred,
I have already read the assertion elsewhere, that if there is no solution with internals, there can't be any with externals.
I have never seen any demonstration. I have seen many examplars where the puzzle solution was much easier with externals (or mixed guardians) than with internals only. But I have not found a couter-example (nor have I searched for one). Oct. 7 Tough puzzle is not a counter-example either, though the solution using 1c9 is not easy to find. Refer to the puzzle thread to read it.
Best regards.