View previous topic :: View next topic |
Author |
Message |
AZ Matt
Joined: 03 Nov 2005 Posts: 63 Location: Hiding under my desk in Phoenix AZ USA
|
Posted: Tue Jun 13, 2006 11:13 pm Post subject: "Very Hard" upgrade??? |
|
|
I have been gone for 9 months (long story); has the "Very Hard" category been updated since I was last here to recognize more difficult puzzles as "Solvable"??? I remember the program having a little difficulty with some advanced techniques, but I thought the two most recent "Very Hards" (the 4th and the 10th) were strong. |
|
Back to top |
|
|
keith
Joined: 19 Sep 2005 Posts: 3355 Location: near Detroit, Michigan, USA
|
Posted: Tue Jun 13, 2006 11:34 pm Post subject: Getting Stronger ... |
|
|
AZ,
I am not sure if you are referring to the solver or the daily puzzle.
Anyway, a few weeks ago the puzzle was upgraded to possibly include X-wings as a technique in the "very hard" category.
Also, the solver was upgraded to solve X-wings.
Keith |
|
Back to top |
|
|
AZ Matt
Joined: 03 Nov 2005 Posts: 63 Location: Hiding under my desk in Phoenix AZ USA
|
Posted: Wed Jun 14, 2006 4:24 pm Post subject: Thanks |
|
|
Now I remember; it was x-wings that gave the old solver trouble, which I assumed went hand in hand with the degree of difficulty in the "Very Hard" category of the daily puzzle. Great upgrade!!! I used to get puzzles from other sites that were rated the highest degree of difficulty, but certified solvable without guessing, and run them on the solver here, and if it came back "unsovable," I knew it was my favorite type of puzzle.
Does it do complex forcing chains now? I think that used to be a problem, too. I used a tricky forcing chains to solve the June 10th puzzle, which was great, but I am not sure that was the only way to solve it (just the earliest way). |
|
Back to top |
|
|
samgj Site Admin
Joined: 17 Jul 2005 Posts: 106 Location: Cambridge
|
Posted: Wed Jun 14, 2006 9:36 pm Post subject: Re: Thanks |
|
|
AZ Matt wrote: | Does it do complex forcing chains now? |
Not yet, and I guess this probably isn't next on the list -- there seems to still be a split over whether the *requirement* to use forcing chains is "A Good Thing" (TM) or not. So I guess I'm not likely to require them just yet (and that's what drives improvements to the solver code), although of course forcing chains may provide good routes to solve puzzles which formally require other steps.
Sam |
|
Back to top |
|
|
AZ Matt
Joined: 03 Nov 2005 Posts: 63 Location: Hiding under my desk in Phoenix AZ USA
|
Posted: Thu Jun 15, 2006 11:26 pm Post subject: Forcing Chains? |
|
|
All right, I have been out of the loop. Is there some philosophical discussion about whether forcing chains is really trial and error in disguise? I beg to differ if that is the argument. The same could be said for colouring. The only art to sudoku is seeing patterns in the numbers and deducing what to try next (and where). |
|
Back to top |
|
|
samgj Site Admin
Joined: 17 Jul 2005 Posts: 106 Location: Cambridge
|
Posted: Fri Jun 16, 2006 10:58 am Post subject: Re: Forcing Chains? |
|
|
AZ Matt wrote: | All right, I have been out of the loop. Is there some philosophical discussion about whether forcing chains is really trial and error in disguise? I beg to differ if that is the argument. The same could be said for colouring. The only art to sudoku is seeing patterns in the numbers and deducing what to try next (and where). |
Well, I don't want to re-open that can of worms. There can be no reasonable argument to say that you shouldn't use anything you like to solve these things, but my inbox suggests that sections of the intended audience consider this at least close to trial and error. Puzzles here therefore don't currently *require* it. I guess I'm trying (not entirely successfully) to tread a line between the need to cater for the masses, and keeping the hard core who have the interesting stuff to say here happy!
Sam |
|
Back to top |
|
|
|