Almost forgot to blog this.
With my SodokuGo program, I can now also generate Sudoku-X puzzles. Posted a few on the NL forum, but not a lot of response. Interesting that you can easily generate 15 and 16 clue sudoku-X. I saw one with 13 clues on a website.
It would be cool to generate my Clueless Specials with SudokuGo.
Tests are going remarkably well. Despite the large number of changes, no errors have been reported. The only problem may be the bandwidth required for the release. With more and more users, the maximum bandwidth could be exceeded with the release of a new version. This version may be OK, but the next…
The following solver techniques have been added and tested:
The ‘Peers’ tool has been added. This tool could be placed between Focus and Markup. It focuses the view to the selected cell and it’s peers. The candidates for the selected cell only are visible and can be maintained. The tool is aimed at people who want to maintain candidates, but do not like the spoiling effect of seeing all the candidates in the entire grid.
Rating system implemented. This rating system must be tested and calibrated. We probably need to add something like mode-change scores, e.g. when pencilmarks are required, or when value patterns need to be examined.
As an option, markup can be combined with the watch and color tools. This combination can be useful for those accustomed to the Simple Sudoku interface. Markup can be toggled on and off independently when watch or color tools are selected.
Fixed a bug that left the previous menu item available when opening single puzzle after a collection. The statusbar panel no longer shows 1/0 or 1/1 when no collection is loaded.
The analyzer has been split into a separate status dialog, with the current situation (filled cells, remaining candidates, etc) and the Analyzer with required solving techniques and ratings.
To avoid confusion with true pencilmarks, pen and pencil have been remaned to Pen 1 and Pen 2. The latter has a dark green color. The erase and confirm functions are still available for pen 2.
New scramble function. Permutate the current puzzle without changing the difficulty. Nice tool to play a variation of an interesting puzzle (which is a whole new puzzle) or avoid copyright claims. Not that the sudoku vault masters would like it (or Wayne G) but this feature will not operate on batches, just single puzzles.
I have decided to introduce a beta testing period, due to the many changes. I want to make sure this all works, before I throw it at the large audience.
So proud this is working!!!
XYZ-Wings are now supported in SudoCue. With the complexity of the program, adding new solving techniques has become more of an administrative task tham simply writing the code to perform the technique. Anyway, this technique is now part of the standard solving techniques suite of SudoCue.
More to come.
I set up a forum for sudoku beginners on a free forum hosting provider. I’m not sure whether or not I’m gonna use it, but there is a gap in the market to embrace and cuddle new sudoku players. With SudoCue I’m at the high-end of the market, but there is a requirement at the lower end.
I have written a few modifications to SudoCue, good enough to release a new version shortly.
The easy methods (naked and hidden singles) have been modified, so that the program no longer does them via the queue but only when really solving the puzzle.
A rating system is in development that has a very fine tuning in the easy range. It emulates the way a human player would search the grid, and hit a possible placement. The score greatly depends on the time needed to search for breakthroughs, modified with bonus and malus for alternative routes, visibility, difficulty of the technique used and stategy bonus.
My clueless monster is well-received. One player spent 6 hours on it, another 3 days. Need a bit more feedback from the intenational audience, but another clueless sudoku is certainly an option.
I read a bit more about the birth of BUG on the players forum. Interesting how such theories develop. I posted a poll on the programmers forum to see what the general opinion regarding uniqueness is. I know what Angus’ opinion is.
What I want to add to SudoCue is a few more tools and tool combinations. The Peers tool should be a nice addition to scan for naked singles without giving them straight away, as with pencilmarks.
Another option should be to see the candidates/pencilmarks with Focus/Watch or Color tool active. It does not hurt, but players may like the way candidates are presented in SS.
I have decided to give it a final go and put all the things I’ve learned about writing a cool Sudoku program into this new version. It will not be a simple upgrade, but a complete rewrite of the program. A lot of posts about the project’s progress are to be expected, but here is the plan:
Is this too ambitious? We’ll see. I’ve contacted the very inactive administrator of the sudoku programmers forum and offered to take over the administration. If that works out OK, then I’ll set up an Open-Source subforum, and this project can be posted there. If not, I will start a new SudoCue2 development forum on the SudoCue.net site. That would be more difficult to plug, but we’ll see…
Not revealing anything yet, but I have registered 4 domains and they’re beauties.
A new application is in the making. Sorry, it is not Sudoku related.
Also, worked on the Sudoku database. Another 10,000 have been added. I am now saving more indirect properties of the puzzles and it is possible to hunt for special combinations of properties. No freebees, tight solving path, breath in solving techniques, large number of eliminations between single steps.
We can now almost supply a Sudoku “made to order”. That would be the real killer in syndication.
Spent some time helping a guy insert the syndicated link to www.SudoCue.net. He finally got it built in and now he’s running a daily Sudoku from my database.
More! More!
Also did some research on the easies in my database. Created a program that does some low-end difficulty analysis. I’m not gonna spill all the beans here, but some of the considerations are:
Refinement in difficulty assessment will help me identify better nightmares.
This is the toughest Sudoku in stock, from my own manufacturing plant. Is it really a complicated one? It just depends on what force is applied against it. A few nice loops could break it… maybe.
 
  Tough is in the eye of the beholder. In this case, Sudo Cue.
Was surprised to see how many people are reading my daily entries here. OK, needed to tidy it up a bit, and translated the most recent posts in English. Blogging is a dangerous activity. Before you know it, everybody is looking over your shoulder.
Finished the Scanning article in www.Sudopedia.org. May need some revisions, but at least it is a consistent story now. Still no cowriters.
I have isolated the 400.000 Sudokus in my Access database that can be classified as ‘easy’. I will run a special program today, that performs a solving speed test on these sudokus. This is something I found out recently. It is not only the requirement of solving techniques, but also the solving speed that determines the difficulty of a sudoku. For the lower grades, this is the most important benchmark. A whole series of sub-grades can be worked out.
But not for me. I have to keep up my reputation, so I will always supply sudokus that are a tad more difficult than the rest of them. This solving speed test will help me pick out the nice ones.