Sudoku Brevity Contest

[Safari]

231?

I'm at 271, don't think I'll get lower with this algorithm.

Edit: Make that 258 :\

[Safari]

230 tokens using the same algorithm I've used from the start

[Safari]

To continue on my monolog

tokxx.exe -a sudoku.cpp
219

What's up with this secrecy? I know more people than BH and Darryl are participating.

Edit: Forgot to delete some unused variables after some recent changes. Tokens = 214.

Edit2: is => are

[Darryl]

To continue on my monolog

tokxx.exe -a sudoku.cpp
219

What's up with this secrecy? I know more people than BH and Darryl is participating.

Edit: Forgot to delete some unused variables after some recent changes. Tokens = 214.

214!!! Good job....I think I need to rework my algorithm

@BH, rand()&random_shuffle() are definitely good enough, but I am not quite I understand your proposal. I believe you are saying that the puzzle generated shouldn't be limited to a certain subset of possible puzzles. While a good proposal, I don't know if it could be easily tested for example like with Safari; if his has a ~32000 max, unless I studied his algorithm, I couldn't easily discern that his has a limit.

Also, BH, in case you didn't realize, string_literal/80 will be integer division (throwing away remainders), so you can have a string up to 159 chars before it's counted as 2

[Darryl]

Well after changing my algorithm and exploring my own heat death solution , I went back to my original algorithm and got it down to 245. Only 31 more to go to catch you Safari

One other thing I thought I'd mention; you must include "int" in front of main() as my new more standards compliant compiler won't allow otherwise.

[Beer Hunter]

I believe you are saying that the puzzle generated shouldn't be limited to a certain subset of possible puzzles.

Right. Ideally, entries should include a quick explanation, though, to save the readers' time.

Also, BH, in case you didn't realize, string_literal/80 will be integer division (throwing away remainders), so you can have a string up to 159 chars before it's counted as 2

So a string less than 80 characters counts as zero tokens? OK, I can use that to reduce my token count. ;-)

EDIT: Are we allowed to use '#include "iostream"', or do we have to use '#include <iostream>'? My current token count is either 174 or 182 (or 175/183, if I add one for having a string 162 characters long).

[Safari]

lol, I thinkI'll give up. I think I can get less than 200 (I'm at 203) but never down to 174

[Darryl]

So a string less than 80 characters counts as zero tokens? OK, I can use that to reduce my token count.

Guess I didn't think it through....however upon more thinking, I think I'll drop/modify the string requirement to no strings over 255 char.

Yes you can do the #include"iostream", however I didn't want to announce it because then everyone will do it, negating the benefit

My goal was to get to 200, but I am already ready to give up too at 240. Safari, did you still want a Sudoku checker... I whipped one up yesterday ( it was actually my failed heat death entry, but the checker part is good, it was the generating part that failed) and I can quickly modifiy it to accept a file as input.

[Alvaro]

My first attempt has 188 tokens, and I haven't even used any obfuscation tricks, but the sudokus generated have a lot more structure than required. Would that be acceptable?

[Darryl]

188!!! Geez, I was proud I just got down to 206!

What do you mean a lot more structure?

The formatting should appear as in the rules, each 3x3 grid is seperated by a space from it's neighbor on left/right and by a linefeed/empty line from it's neighbor above or below it.

[Alvaro]

No, I am not talking about the formatting. What I mean is that the sudokus generated have strange properties, like 1, 2 and 3 always appearing on the same column of the top-left 3x3 square.

Oh, I had forgotten about the formatting. With an improved generator (not so obvious structures) and the correct formatting, I am at 209, so don't panic just yet.

[Darryl]

I kind of figured after I posted you were talking about number arrangment/patterns or such. I don't have any problems as long as it generates a "sufficient" variety and randomness, but BH, might take exceptions, opting for solutions that don't use just a subset of the whole, but can theoretically generate any sudoku.

Mine can generate any, and i am guessing BH's can, but I think Safari's is only a subset generator.

[Darryl]

lol, I thinkI'll give up. I think I can get less than 200 (I'm at 203) but never down to 174

Don't give up yet, i just hit 203 still using my original algorithm.

[Safari]

but I think Safari's is only a subset generator.

Are you talking about my 186 token solution?
















Okay, that solution works theoretically not yet practically. I'm having a hard time accesing the array at the right places using 1 single loop + 2D-array (i/3*3+s%3 etc etc, it actually gave me a headache).

[Darryl]

but I think Safari's is only a subset generator.

Are you talking about my 186 token solution?

Yawn, 186 is soooo yesterday
let me know if you reach 178