All puzzle types
All Sudoku types
Odd Pair Sudoku
Killer Sudoku Pro
All puzzle states
Started (no save)
15 per page
30 per page
50 per page
100 per page
All on one page
Click on column headings to sort.
Back to the forum
I've just updated the Killer Sudoku cage code and now there should hopefully no longer be any instances where the dashed line cages don't quite line up with themselves between cells. So hopefully the occasional artefacts with these puzzles are now gone for good. I had fixed the 'phantom lines' that sometimes appeared a month or so ago, but with today's change I hope the last of the occasional minor display glitches is gone.
And what's more, I've fixed a very long-standing issue whereby if you had the mouse button pressed down as you wandered out of the window and then came back into the window, while playing Slitherlink or Hanjie or any other click-to-shade puzzle, you would carry on shading even if you'd long since released the button.
This means there's some new mouse handlers in the player that provide this top-level support to prevent the cases when the mouse button gets "stuck down". I am only able right now to test this on current browser versions, but due to the way it's implemented I have no reason to think it won't work even on much older browsers. If you spot any problems do please let me know.
1) The 'average time when solved with no aids or only minor aids' statistic is back on the comments page, after it went missing when I changed to the 'without' and 'with' aids system of displaying stats.
2) On the comments page you now get a small purple badge (well okay, square) next to your username if you're a daily puzzle subscriber. Over time I plan to add more badges so that if you for example solve a lot of puzzles you'll get a badge for that, and so on - and I'll probably make them actually badge-like at that point. :)
3) Other minor fixes. For example if you subscribe then the exact reason why you can't play an old puzzle pre-dating your subscription is now more accurately described on the comments page.
Meanwhile there've been no shouts of horror since I rewrote the way the display and mouse handling works, so I'm guessing this is working for everyone. In particular the font sizes are now even more consistent and the solution numbers in sudoku et al are now notably larger within their cells which helps for a start on small devices like the iPhone.
I'm aware if you complete a puzzle on the iPhone the browser has an issue letting you rate the puzzle for difficulty - I'm not sure why but I'll fix it. However the actual puzzle play works perfectly, particularly on iOS6 where in landscape mode you can hit the "full screen" button and it becomes just that.
I'm working on proper touch support for Slitherlink (which I'll then extend to the other shading-based players) at the moment, and it's turning out to be more complex than I imagined. I don't want to use any simulated mouse events because these can be slow to trigger and have a whole host of other issues, so for what the players need to do I essentially have to use HTML touch events. The problem is that these are designed for returning coordinates rather than objects, once you start the touch movement (or drag, in other words). There is an event called 'touchenter' but it doesn't fire on all browsers, including iOS, so it's no use.
You can however convert coordinates to objects using a browser call, but there's just one issue. It behaves differently on newer and older iOS versions, and differently again on the Nexus 7 or indeed any Android Google Chrome browser (where it picks the wrong object). So I'm now having to write a set of patches to pick up all the different versions, and that's just on the browsers I can test. It's like the early days of web development again, where nothing worked consistently across browsers.
I can be pretty certain that whatever I write now will need fixing for a new set of browsers very soon, but c'est la vie!
PS: And it turns out the Google Chrome fix doesn't work in portrait mode, only landscape. Possibly an issue with the size of browser decorations. It works fine in Firefox on Android and, I believe, the Android browser (but I don't have it to test). Looks like at least for Nexus 7 users I'll just have to leave it and wait for Google to fix it, or suggest they install Firefox. :)
I've also updated the front page of the site and the puzzle information pages linked from it. So far I've rewritten half of the information pages, to be more precise, and hope to update the rest within the next 24 hours. However this is mostly just to improve the experience for new users and isn't so relevant to existing users.
Meanwhile, I've now given the site a general polish all over. There are shinier-looking buttons and sparklier forms on many pages.The tabs at the top of the screen now shade in pale blue when you're on the page they refer to too, which hopefully makes navigation a little easier as well.
I've also fixed the odd layout issue on the right-hand panel in the player that lingered after I did the work to update for touch devices. In the process I made the buttons a bit larger too which meant I needed to free up a little space, so I've compacted the pencilmark options and removed the "Add all" button which added every pencilmark to every cell, which I can't imagine anyone ever used. I never have. So that's now gone and I moved six lines (including spacing) into two as a result, reducing clutter in the process too I hope.
If you spot any problems please do let me know. The only player bugs I'm aware of affect touch: you have to hit undo/redo several times sometimes in the shading players because it adds multiple events for the same actual happening to the undo buffer (I broke this when I fixed support for computers with touch and mouse and since no one's complained it hasn't been a priority to fix). And if playing Hanjie and you tap outside the puzzle but inside the play area then it toggles the last placement you made which of course it shouldn't do. Again this is due to support for both touch and mouse and I haven't prioritised it because it's an unusual problem that no one seems to have.
THANK YOU for commenting so quickly!
I've also updated pencilmark placement in all of the players, so nothing should any longer be obscured at the edges or e.g. by a consecutive bar. This does mean on the consecutive/XV/odd-pair/kropki/killer/killer puzzles that the pencilmarks are quite close together at the existing font size. I'd be interested if you think this is okay or the pencilmarks should be a little smaller when viewing those puzzles.
While doing this I also added support for negative clues such as "-3-" in Killer/Calcudoku. For example -3- could be solved with 4, 5, 6, since 6-5-4 = -3.
As always, if you spot any problems please let me know.
I've also made some changes to the 'dynamic total' and 'clue fits' code for 9x9 Killer to support "Killer Zero" puzzles. This change should be transparent but if you spot any problems with this function please let me know. The player now will handle Calcudoku Zero and Killer Zero puzzles, which I'll post examples of in the near future.
Finally, I fixed a minor bug there's been for a while where if you pressed 'give up' on certain sizes of Killer/Calcudoku puzzle you'd find the the green "incorrect" notes weren't always in sensible positions. Now they always form a rectangle around the correct digit. The green digit shows the value you had previously entered.
Oh, and the site now supports 5x5 and 7x7 Calcudoku too, which it didn't previously.
Back to the forum