Page 18 of 28
Posted: Thu 28 Apr, 2011 12:33 am
Posted: Thu 28 Apr, 2011 8:06 am
On a different point, how on earth can anyone fill in all but one of the squares in 15 seconds or so. I can't remember the fastest time exactly. It took me 45 seconds and I don't think hardly any of that was thinking time. Very frustrating as I don't think I could improve my times.
Posted: Thu 28 Apr, 2011 8:23 am
david mcneill wrote:On a different point, how on earth can anyone fill in all but one of the squares in 15 seconds or so. I can't remember the fastest time exactly. It took me 45 seconds and I don't think hardly any of that was thinking time. Very frustrating as I don't think I could improve my times.
David, are you using a combination of mouse and keyboard? ie, roll mouse over cells, while tapping the digits on the RHS keypad?
Also, Ctrl-Return submits quicker than using the mouse to click Submit. Practice on this one
. I can solve it in around 30s.
You can see previous highscore tables by entering the date into the highscore table: Mars did a 13s clean, Melon did 15s with 1 FVS.
On a side-note, I've asked Berni about the possiblility of maintaining an archive. He's yet to respond, which tells me he's thinking about it.
Yeah, I missed that '2' also.
Posted: Thu 28 Apr, 2011 10:50 am
yeah, the croco interface is very good for inputting numbers in that you don't need to click the cell in question, you only need your mouse to be hovering over it. This means that the fastest times are people literally sweeping the mouse over the puzzle and synchronising their keystrokes. The fastest I could manage was 18s though - and in that sort of context 5 seconds is a massive amount of time. You'd have to ask Thomas/Palmer what they do, although I'm certain they are both more than nifty with a keyboard...
Posted: Thu 28 Apr, 2011 2:54 pm
My mouse control has always been rubbish (certainly wasted 5 seconds with incorrect connections on the Masyu today, which seems more sensitive than it used to be in a bad way). So I'm just using my two hands to hit the arrow keys and the numpad and basically typing as fast as I can. Since I don't use the mouse, learning control + enter saved me a second this time so I beat pwahs instead of tying him. In the past I would have to regrab the mouse and find the button.
Posted: Fri 29 Apr, 2011 8:16 am
That's interesting - Jason always said that using arrow keys for kakuro on nikoli was a lot quicker for him, but I've always preferred using a mouse.
On an own-trumpet-blowing note, I've finally managed to crack the top 100, after stagnating for a long while last month. We'll see how long that manages to last!
Posted: Fri 29 Apr, 2011 8:52 am
I always use mouse and keyboard, but I do have a real problem with the Masyu interface with the mouse straying to the wrong line e.g. selecting a horizontal instead of a vertical. Is that my mouse, or do other people have that problem?
Posted: Fri 29 Apr, 2011 11:55 am
grr. a perfectly solved masyu in a reasonable time, but a rogue line had appeared.
Posted: Fri 29 Apr, 2011 1:06 pm
I had that problem the first masyu this week (I think my mouse cable was a little tangled), but obviously not on the current one. What I find helped was increasing the size of the puzzle applet (as has been described earlier in this thread somewhere) - the idea being is that bigger cells means you have bigger margins of error when you are dragging the line out.
The other thing with the masyu is to almost ignore what the cursor is highlighting and dragging out the loop anyway. That always seems to work for me.
Posted: Fri 29 Apr, 2011 2:08 pm
am finding the masyu interface far more difficult since it's changed.
I had learned to click and drag in a particular way - but that now seems to put lines not quite where I was expecting.
similar to what drsteve and motris describe.
Posted: Fri 29 Apr, 2011 3:11 pm
The main mode of failure for me is dragging a line to the center of a box and, because I went a few pixels past center, getting an extra unwanted segment at the same time. This makes me now have to be extra careful when drawing so that I don't have a hidden segment, say, near two black circles to Fehlversuch. I'm already at a somewhat larger window, but I guess I can see fully expanding the size when it's a Masyu puzzle and restoring to my intermediate size for all other puzzles.
Posted: Sat 30 Apr, 2011 7:03 pm
I'm still not happy with the masyu interface. The problem lies in the difference people look at the puzzle. Some look at the cells while others look at the segments. Dragging programmed for eighter way leads to problems in the other way. I tried to find some sort of compromise for the time being. A better way would be, that people can select their prefered way beforehand (and maybe even their own keys), but that means a lot of work on the applets which I can't spend the next time.
Concerning the questions of Alan towards a croco forum: There has been an official croco forum years ago, but almost noone used it. Then the logic masters forum rose and people started to discuss croco topics there so I decieded to close down that older forum. I'm pretty happy with people using the forum at logic masters. This lead to good fluctuation between both sides and is in my eyes a win-win-situation. From my point of view, it look like this forum goes to be the english pendant of the logic masters forum and I'm pretty happy with that.
Alan also asked me to open an archive with old Ü-Rätsel. This idea has already been addressed several times to me and I gave it a lot of thought, but I still think, that it's better not to have such an archiv. First of all, the Ü-Rätsel are all computer generated and I don't think, that such puzzles deserve to be archieved (although I collected them all since 2005). I also fear that so much puzzles make the Ü-Rätsel (and may be other sites with puzzles) less attractive and last, I hope that poeple might buy the denksel magazin (a pdf file) from time to time, which contains over 100 puzzles (all but one special puzzle are computer generated).
Posted: Sun 01 May, 2011 6:13 pm
Berni, I agree that archiving all Ü-Rätsel would lead to a puzzle inflation on the croco-puzzle site. But what about archiving only Ü-Rätsel that are special in some way? For example puzzles where more than 50% of competitors had a Fehlversuch, puzzles that were only solved by less than 50% of the people who attempted to solve it, or puzzles that took the fastest solver over an hour? Alternatively you could add such puzzles to the puzzle portal, of course... (or somebody else could if you give them the permission to publish the puzzle
One note on the new Masyu interface: I am not very good at solving Masyu, but the new mouse behaviour works like a charm for me.
Posted: Wed 04 May, 2011 8:54 pm
Congratulations to afcwarren on achieving 20th kyu! I look forward to Warren and Liane overtaking me before long, but I'm going to try not to make it too easy for them...
Now I wonder whether there are other not-yet-belted names who actually represent British players without me being able to recognise them?
Posted: Thu 05 May, 2011 11:19 am
Thanks Dickoon. Even if you didn't play at all, I think I'm at least 2 months behind you in the Ratingliste so there's no need to worry!
Posted: Thu 05 May, 2011 8:02 pm
Now I wonder whether there are other not-yet-belted names who actually represent British players without me being able to recognise them?[/quote]
My son joined up a while ago, but he's not played for a while. I wonder whether you'd be able to recognise his name in there?
He's a Sheffield Wednesday fan
Posted: Fri 06 May, 2011 7:17 am
Based on that clue, would it be
sadowl, from the club's logo
Posted: Fri 06 May, 2011 1:53 pm
You're right! I don't suppose there'd be too many points for that (probably even fewer than they normally score in a season
Posted: Fri 13 May, 2011 8:05 pm
Congratulations to ronaldx on being the first modern era UK puzzler to make it to 9th kyu!
I think there are three big promotions, where it's very clear you've been promoted because the shape of the numbers and letters after your name changes. Going from having nothing to having something is a big change. Going from having a small k to a big D is a big change. Going from having two digits to one digit is the other big change. Having any sort of status shows dedication; a single-digit status shows talent as well as dedication, and the big D shows you're something rather special.
Sometimes I wonder if there are other big promotions as well and, you know, I'm not sure there are to the same extent. It's fascinating to see how high a belt that the not-just-world-class-but-world-leading solvers can attain, but the only - "only"! - thing special about 7th dan is that it represents the highest level yet, and probably ever, attained, rather than there being a particular property of the number 7. Your 5k and your 15k might be logical intermediate points between the big promotions. Speaking of which, rodders has achieved the latter today as well!
Posted: Fri 13 May, 2011 8:20 pm
This seemed like a more relevant place to post a walkthrough to yesterday's inequality sudoku:
The main starting point is C5. Without too much though you can immediately place:
R1C5 = 2
R5C5 = 8
R6C5 = 6
R7C5 = 1
R8C5 = 4
The next thing to notice is the chain of 4 inequalities towards the top left. This gives R3C3 = 8. Further, looking down C2, it emerges that the only place for an 8 is at R6C2.
There's more joy to be had in C2. With a bit of head-scratching (i.e. considering the placement of the crossing 1 and 9) you can see that there are only 4 numbers that can be placed in the chain from R7C2-R9C2. These are 2,3,5 and 7. In particular, note that the only candidates for R9C2 are 5,7.
Now we note that the only two candidates for R9C6 are 5,7. We have ourselves a pair. In particular, this means that R9C4 = 3.
Now lets look at R6. The way the inequalities are going means that there's only one place the 1 can go; R6C8=1 R6C7=1. This in turn leads to:
R9C8 = 1
R8C1 = 1
R7C1 = 8
R8C4 = 8
R6 has more fruit to bear. Notice that the 2 can only go in C1 or C3. It can;t be C1 because of the 1 already in that column, hence we have
R6C3 = 2
R5C3 = 1
R3C2 = 1
Further, the 2 can only go in one place in C2, giving:
R7C2 = 2
R9C1 = 6
R9C7 = 2
R4C3 = 6
R2C2 = 5 R2C3 = 5.
The 9 can now only go in one place in R7:
R7C3 = 9
R8C3 = 7
R1C3 = 3
R8C2 = 3
R9C2 = 5
R9C6 = 7
R7C4 = 5
R4C2 = 4
R6C8 = 4
R7C8 = 3
R5C8 = 2
R4C4 = 2
R4C6 = 1
R5C6 = 4
R1C6 = 5
R2C6 = 8
A nice little flurry, I'm sure you'll agree. The puzzle is nearly cracked, but there's a couple more things to notice. The chain of 3 inequalities in C9 can only be satisfied by:
R4C9 = 3
R3C9 = 2, leading to
R2C1 = 2
R3C1 = 4
(R4C3 = 6)
R3C7 = 3, as well as
R5C1 = 5 R5C1 = 3
R6C1 = 5.
The finishing blow is given by noticing the 79 pair in R1C1 and R1C2 means that R1C8 = 8. The rest I'd probably just be insulting your intelligence to spell out!
Posted: Fri 13 May, 2011 11:32 pm
I rarely throw puzzles in the bin, but this is twice I've had to pull this one out now - Lesson learned! ...the stack keeps growing...
I was right with you until:
detuned wrote:Now lets look at R6. The way the inequalities are going means that there's only one place the 1 can go; R6C8=1. This in turn leads to:
R9C8 = 1
R8C1 = 1
First, clearly a forgivable typo - R6C7
is clearly a 1. However, why can't R9C1 be a 1 for the next step? *thinks*... Ahhhh, well, now you point it out, the TR 3x3 has to have its 1 in Col 9 - THAT's why... and would have saved me around 3 hours if I'd spotted it. doh! ...reading on...
The chain of 3 inequalities in C9 can only be satisfied by:
R4C9 = 3
Stumped on this for a second until I saw the 7/9 pair in C2/4 -> R5C9 = 5/6 -> R4C9=3 ...reading on...
Great work Tom! How anyone saw all that in under 30 minutes is quite incredible!
Not too sure about your cell naming convention. Usually for walkthroughs we use cols A-I and rows 1-9, so your R5C5 would just be E5.
Hope you don't mind me clearing up the typos. If you have a copy of the puzzle image, it might be worth adding to your post so the walkthrough will actually mean anything to anyone else after today.
Posted: Sat 14 May, 2011 8:08 am
The whole RXCY seems to be a fairly conventional thing for sudoku, although obvious seems a little more cumbersome now you put it like that. I have no idea why it evolved like that - never stopped to think about it.
Also, thanks for pointing out the typos - perhaps you can see why I mess up so many online puzzle contests
! I was solving a screenshot in paint and going back and forth between windows when typing it out. I'll go edit in the screenshot now...
Posted: Sat 14 May, 2011 12:28 pm
detuned wrote:The whole RXCY seems to be a fairly conventional thing for sudoku,...
I've no idea about conventions but I do know that I definitely prefer the RXCY. With the alternative I always have to double check if the letters are for rows or columns and which letter corresponds with which number. So RYCY is much
faster for me to read.
Posted: Sat 14 May, 2011 7:40 pm
R3C5 is much easier understood for me too.
Posted: Sat 14 May, 2011 8:55 pm
I suspect chess players prefer E5 or similar, while other puzzlers prefer R5C5