« homeBrowsing all comments (spoilers: on | off) (info: on | off) next 25
Exit 1 Sun 08 Dec 2013 14:46
<mark> 'Exit 1' uploaded by mark:
As promised, cook away.
 
Mods chat level Sun 08 Dec 2013 14:43
<mark> (SPOILER) Seems you have been busy Dave and after reading your earlier comments i've decided to make Exit 1.
I had a third solution to it so i hope my mod forces it.
I really liked this level and gave it 10 for style first time around.
Also read your comments N7 and you might also be tinkering with this hope this does not stop you.
Have fun level coming soon.
 
No Three in a Line (9x9) Sun 08 Dec 2013 10:44
<mark> (SPOILER) Added speedrun: 385 moves (old: 578).
Ctrl+0 did not grab speedrun.So here is a slightly diifferent solution same nett result.
 
Mods chat level Sun 08 Dec 2013 04:55
<Dave> (SPOILER) N7. Exit one is worthy of a level in its own right. To be honest, I haven't got very far with my plan because I've been somewhat sidetracked by other level ideas for a while. A mod won't spoil the idea, I'll just add it to the inventory of things to be deployed later. Please do not hold back on my account.

You already know what needs to be done to get exit one ... The fastest option (thus far) uses 2&3 (2 traps 3) but there are other ways: my first solve used 3&8.

Xeno has a new one based on a old Milton level, should be up very soon. I have other levels in the pipeline but nothing even close to ready so don't hold your breath.
 
Level comments: escape.spacebar.org/f/a/escape/level/3724 Sun 08 Dec 2013 00:29
<N7DOT> I still am unable to get to exit number 1... Pesky bots! I'm wondering though, Dave, if uploading a modification attempting to force escaping through the upper exit using minimalistic and 'just barely impossible' fixes to block the other doors (and removing the deadly aspect of the laser maze) would spoil the version you were intending to upload. If not, then once I can *actually* get to door 1 maybe I'll upload such a modification and call it 'The Accidental Puzzle' or something like that.
 
No Three in a Line (9x9) Sat 07 Dec 2013 22:51
<Dave> 'No Three in a Line (9x9)' uploaded by Dave:
9x9 part 3 (fixed).
 
No Three in a Line (9x9) Sat 07 Dec 2013 22:49
<Dave> Moved to graveyard:
Oh dear, that was silly. Fix coming... Mark, you can ctrl-0 to get speedrun back :)
 
Force of Attraction (Live) Sat 07 Dec 2013 19:13
<devin> I won by accident.
 
No Three in a Line (9x9) Sat 07 Dec 2013 12:02
<mark> (SPOILER) Added speedrun: 385 moves (old: 456).
 
Mods chat level Sat 07 Dec 2013 04:23
<Dave> OK, I'll have a quick try at explaining the transponder version because I'm never going to upload it anyway. I never fully tested it because it wasn't good enough but ...

You have a grid full of panles with crossovers in bizzaro and park transponders on it. The panels map to an identical grid of remotes with blocks in bizzaro. The remotes map back to the same spots on the original grid. If the bizzaro grid is fired horizontally from the right from it will turn all the crossovers making a path to the rightmost transponder, repeat from the left. Fire across the original grid, any number other than 2 or certain cases of higher even numbers will fail, 0 cancels out. Higher evens than 2 can be ignored anyway. Refire the same way to reset the grid to its original state and repeat vertically. Hope this makes sense ...
<Dave> PPS. I am still clueless about your dalek/teleport direction reversal mechanism. Every time I tried I came unstuck with multiple consecutive reversals.
<Dave> N7. Continuing from previous comment outside of the level comments ...

I'm curious, you are full of interesting ideas, how far did you get with this before? The difficult bit is testing two axes using only one 'singularly active' piece without introducing a huge maze of remotes. (by 'singularly active' I mean a piece that can only switch over one other cell). In the transponder version I mentioned before, the transponders are 'multiply active' because they are used as both a trigger and part of the test procedure allowing a bi-axial test for exactly two pieces on each line on both axes, this makes the zero test redundant.

The final implementation using 'singularly active' pieces (ie. spheres) only checks for an even number on each axis but is backed up by a 'fail on zero' test. This only works becuse four in a line is not possible without failing either the 'zero test' or the 'parity test' somewhere else.

PS. If you solve these levels every dalek dies, if you fail they don't. If you have access to a large enough monitor (I only have this at work), the animation of the test procedure, especially for Queens, is really good. :)
 
No Three in a Line (9x9) Fri 06 Dec 2013 23:37
<Dave> (SPOILER) Hi N7.

I found the easist and most compact way to implement this would be to use transponders as the pieces. It's rather hard to explain it in a few sentences though. The problem is: the underlying puzzle, although mathematically interesting, is way to easy to make a decent Escape level. So I ditched the idea completely and set about making it work with balls. This way it would not be too hard but make a really interesting speedrun challenge. Although the final parity checking mechanism is really simple, it did take me a while to come up with it.

Does this answer your question?
 
revenge of the malformed levels 4 Fri 06 Dec 2013 21:56
<N7DOT> woo solved it :) (with 2007 version of course)
 
No Three in a Line (9x9) Fri 06 Dec 2013 20:25
<N7DOT> Oops, forgot a question I had: Is there a good reason why spheres were used instead of steel?
<N7DOT> You beat me to it! I had been planning an implementation of No Three in a Row, but that didn't get much work on it.
<Jim> (SPOILER) Added speedrun: 456 moves (old: 478).
<mark> (SPOILER) Added speedrun: 478 moves (old: 578).
Nice puzzles Dave.This one justed clicked on the first thing i did.
N queens (9x9)took me over an hour.
<Dave> The rules: Every sphere must be in the 9x9 field, no more than two in any row or column. Simple as that.
<Dave> 'No Three in a Line (9x9)' uploaded by Dave:
9x9 part 3.
 
Queens (9x9) Fri 06 Dec 2013 01:32
<Dave> 'Queens (9x9)' uploaded by Dave:
9x9 part 2.
 
Mods chat level Fri 06 Dec 2013 01:31
<Dave> (SPOILER) I've got a couple more levels for the 9x9 series. Nine queens is pretty simple, it was just fun to implement with the self-imposed restriction of an orthagonal d-map.
 
Rook's Tour (9x9) Tue 03 Dec 2013 23:07
<Jim> (SPOILER) Added solution "Untitled":
Not much different really but I uploaded it anyway so you could see Dave.
 
Mods chat level Mon 02 Dec 2013 22:16
<Dave> (SPOILER) Mark, after you suggested 9x9 I just decided to build the level and hope I would be able to figure out a way of solving it afterwards.

Never been much of a chess player, I've always prefered games with a mixture of skill and chance.

Jim, I think you've nailed it with Improv 2.5. I don't think I'll be taking it any further either.
 
Not difficult Mon 02 Dec 2013 20:39
<Dave> (SPOILER) Added speedrun: 52 moves (old: 53).
 
« homeBrowsing all comments (spoilers: on | off) (info: on | off) next 25