The New Sudoku Players' Forum

[SteveG48]

Another from the local newspaper:

Code: Select all

 *-----------*
 |.1.|..6|.5.|
 |3..|...|...|
 |678|9..|2.4|
 |---+---+---|
 |...|4..|..7|
 |..4|.1.|6..|
 |7..|..3|...|
 |---+---+---|
 |8.9|..5|172|
 |...|...|..9|
 |.4.|2..|.8.|
 *-----------*


[Cenoman]

Code: Select all

 +------------------------+---------------------+--------------------+
 |  4      1       2      |  37   37     6      |  9     5     8     |
 |  3      9       5      |  8    2      4      |  7     16   c16    |
 |  6      7       8      |  9    5      1      |  2     3     4     |
 +------------------------+---------------------+--------------------+
 |  1259   23568   136    |  4    689    289    |  358   129   7     |
 |  259    2358    4      |  57   1      2789   |  6     29    35    |
 |  7      28     e1-6    | a56*  89     3      |  48    249  d15    |
 +------------------------+---------------------+--------------------+
 |  8      36      9      | a36*  4      5      |  1     7     2     |
 |  25     2356    367    |  1    3678   78     |  345   46    9     |
 |  15     4      a1367*  |  2   a3679*  79     |  35    8    b356   |
 +------------------------+---------------------+--------------------+

Almost kite (*):
[(6)r6c4 = r7c4 - r9c5 = r9c3] = r9c9 - (6=1)r2c9 - r6c9 = (1)r6c3 => -6 r6c3; ste

[eleven]

Very similar:

Code: Select all

 *-------------------------------------------------------------------*
 |  4      1       2      |  37   37     6      |  9     5     8     |
 |  3      9       5      |  8    2      4      |  7     16   e16    |
 |  6      7       8      |  9    5      1      |  2     3     4     |
 |------------------------+---------------------+--------------------|
 |  1259   23568   136    |  4    689    289    |  358   129   7     |
 |  259    2358    4      |  57   1      2789   |  6     29    35    |
 |  7      28     b16     | a6-5  89     3      |  48    249  e15    |
 |------------------------+---------------------+--------------------|
 |  8      36      9      | b36   4      5      |  1     7     2     |
 |  25     2356    367    |  1    3678   78     |  345   46    9     |
 |  15     4      c1367   |  2   c3679   79     |  35    8    d356   |
 *-------------------------------------------------------------------*

6r6c4 = 6r6c3 & r7c4 - r9c35 = r9c9 - (6=15)r26c9 => -5r6c4

[SpAce]

Short and elegant.

Code: Select all

.-------------------.---------------------.--------------------.
| 4     1      2    |  e37   37      6    |  9     5       8   |
| 3     9      5    |   8    2       4    |  7     16     b16  |
| 6     7      8    |   9    5       1    |  2     3       4   |
:-------------------+---------------------+--------------------:
| 1259  23568  136  |   4    689     289  |  358  b129     7   |
| 259   2358   4    |  d57   1      e2789 |  6    b29     d35  |
| 7     28    b16   | bd56  a89      3    |  48   b249    c15  |
:-------------------+---------------------+--------------------:
| 8     36     9    | ae36   4       5    |  1     7       2   |
| 25    2356   367  |   1    3678    78   |  345   46      9   |
| 15    4      1367 |   2    7-369  e79   | a35    8   abde356 |
'-------------------'---------------------'--------------------'

(35)r9c79.(6)r7c4.(9)r6c5 = (61)r92c9|r6c43|(921)r654c8 - (1=5)r6c9 - (57)r65c4|(53)r59c9 = (73)r17c4.(6)r9c9.(7,9)r59c6 => -369 r9c5; stte

Or my variant of the common theme.

Code: Select all

.--------------------.------------------.----------------.
| 4     1      2     |  37  37     6    | 9    5     8   |
| 3     9      5     |  8   2      4    | 7    16    16  |
| 6     7      8     |  9   5      1    | 2    3     4   |
:--------------------+------------------+----------------:
| 1259  23568  136   |  4   689    289  | 358  129   7   |
| 259   2358   4     |  57  1      2789 | 6    29   a35  |
| 7     28    b16    | b56  89     3    | 48   249  a15  |
:--------------------+------------------+----------------:
| 8     36     9     | b36  4      5    | 1    7     2   |
| 25    2356   367   |  1   3678   78   | 345  46    9   |
| 15    4      137-6 |  2   379-6  79   | 35   8    a356 |
'--------------------'------------------'----------------'

(635)r956c9 = (56)r6c43,r67c4 => -6 r9c35; stte

(The comma between cells replaces '&' in my new notation under testing. Thus that reads 56 in both r6c43 AND r67c4.)

[SteveG48]

(635)r956c9 = (56)r6c43,r67c4 => -6 r9c35; stte

(The comma between cells replaces '&' in my new notation under testing. Thus that reads 56 in both r6c43 AND r67c4.)

SpAce, I don't like this one. To me, (56)r6c43,r67c4 means that both 5 and 6 are true in the set of four cells r6c43 and r67c4. You want to read it as both 5 and 6 being true in each of two different pairs of cells. You may be saving a few keystrokes over (56)r6c43&(56)r67c4, but at the cost of introducing ambiguity.

[SpAce]

Hi Steve,

(635)r956c9 = (56)r6c43,r67c4 => -6 r9c35; stte

(The comma between cells replaces '&' in my new notation under testing. Thus that reads 56 in both r6c43 AND r67c4.)

SpAce, I don't like this one.

That was a totally expected reaction. The only surprising thing is that it's a bit late! I've mentioned this idea a couple of times already and asked for feedback, but no one has said anything.

To me, (56)r6c43,r67c4 means that both 5 and 6 are true in the set of four cells r6c43 and r67c4.

Yes, that is the standard Eureka interpretation. Like I've said before, I'm not even attempting to change that in Eureka. I'm just testing it with Eureka to see whether it works in practice or if I need to figure out another way to do it in "Neureka".

You want to read it as both 5 and 6 being true in each of two different pairs of cells.

Yes.

You may be saving a few keystrokes over (56)r6c43&(56)r67c4

It has nothing (or not much) to do with the number of keystrokes. It would be perfectly correct and standard if written (56)r6c43&r67c4, so there's no difference in length if written that way. The real issue is the '&' which makes everything unreadable unless surrounded by white space. That's why I want to get rid of it. I've always hated the '&' for that reason, even though it's the most intuitive symbol for AND.

To make it readable with the '&', it would have to be written ((56)r6c43 & (56)r67c4) or (56r6c43 & 56r67c4) or 56(r6c43 & r67c4) or (56)(r6c43 & r67c4). I don't like any of them, though they're definitely better than without the white space. (Note that eleven has apparently seen the same problem, since he often uses white space around the '&', though not in the way I would.)

Edit: Actually, the (56)r6c43&(56)r67c4 is not that bad because the parentheses improve readability even with the '&'. So, in this case you're right that it has something to do with keystrokes, because the biggest problem is with the short form. (56)r6c43&r67c4 and 56r6c43&r67c4 would be really bad, and 56r6c43&56r67c4 only slightly better (or not). (56)r6c43 & r67c4 or 56r6c43 & r67c4 would be more readable but I don't like them (especially the latter) either, because it's not immediately obvious that 56 applies to both sets of cells, and unbracketed white space between links doesn't look good to me.

Even with the current standards the '&' could be avoided by using the comma but it requires duplicating the digits: 56r6c43,56r67c4 or (56)r6c43,(56)r67c4. Both of those are correct and readable without any white space. What I want is to make the comma work logically the same way even if the digits aren't duplicated.

but at the cost of introducing ambiguity.

There's no ambiguity in the new language for which I'm testing this. It looks so different from Eureka anyway that there's no room for confusion. It will have its own, well-defined syntax rules. Whether this or something else will replace '&', I'm not sure, but I'm making no compromises on readability. I've always hated the '&' so it will definitely be banned in any situation that doesn't warrant white space in the first place.

I've always hated the standard meaning of the comma too, so this looks like an opportunity to get rid of both annoying features (by also adding the dot). But, I'm not yet sure of the exact roles of the dot and the comma. It's hard to achieve semantic consistency when the same symbols are used between digits and cells in slightly different roles. We'll see. Like I've said, I'm open for feedback.

[SteveG48]

That was a totally expected reaction. The only surprising thing is that it's a bit late! I've mentioned this idea a couple of times already and asked for feedback, but no one has said anything.

Been kinda busy lately . I read as much as I can.

[SteveG48]

There's no ambiguity in the new language for which I'm testing this. It looks so different from Eureka anyway that there's no room for confusion.

So I assume that you'll be posting a tutorial on your new language sometime soon?

Sooner rather than later would be good. I would be interested in seeing you post your solutions in both standard Eureka and the new language rather than testing the new ideas using (nominally) Eureka. The rest of us could try it as well. The tutorial needn't be perfect the first time out; build on it as time goes by.

[SpAce]

Hi Steve,

So I assume that you'll be posting a tutorial on your new language sometime soon?

Sooner rather than later would be good. I would be interested in seeing you post your solutions in both standard Eureka and the new language rather than testing the new ideas using (nominally) Eureka. The rest of us could try it as well. The tutorial needn't be perfect the first time out; build on it as time goes by.

Thanks for expressing interest. I understand your point, and I'd like to do exactly that. However, some of my design choices (mainly the symbols and their exact roles and definitions) are still fluctuating so much that I'm not sure if it's a good idea to confuse people with the full set yet, as many things might change. Of course it would give a chance to get feedback and make better choices based on that. Another thing is that I thought it would be a good idea to have a language reference for Eureka first, so it could be referenced to show what exactly is different.

One part that is pretty much finalized is the 'Alien Coordinates' aka 19aIrZ, which is the native coordinate system in 'Alien Runes' aka Neureka. The only thing fluctuating in that is if and how it could be extended to support some variants, but that doesn't change anything about how it's used with the vanilla sudokus. I presume it's also the hardest part to get used to, so perhaps I'll start posting that in parallel with the normal rYcX-coordinates. It's a separate component that can be used with any sudoku language, and vice versa, Neureka can use the normal rYcZ/bBpP coordinates too.

You can find my description of 19aIrZ here. Feedback is welcome.

Added. An example.