|
Pages: [1] 2 3 ... 13
|
 |
|
Author
|
Topic: Werbad's Puzzles (Read 15266 times)
|
|
Werbad
|
|
|
|
|
« Last Edit: June 17, 2008, 10:08:25 AM by Werbad »
|
Logged
|
|
|
|
|
Bucky
|
Inspection Line( joducuv): bocyboh. I'm still curious as to how you intended this to be solved.
|
|
|
|
|
Logged
|
That is the most ingenious method of solving an impossible puzzle that I have ever seen.
|
|
|
|
|
|
Werbad
|
Inspection Line ( joducuv): buvatubStreams ( rilacog): jypelutEncryption Hash ( tukafur): punytodWhen I had finished my solution for Change i thought of how evil it would be to remove all barrels but the 1 barrel, thus resulting in the poor version... Since I wasn't sure if you would have enough room I didn't bother solving it... Nice work on that one. Made some new ones: Messenger gegygifRotation vagosuv
|
|
|
|
« Last Edit: February 19, 2008, 02:09:12 PM by Werbad »
|
Logged
|
|
|
|
|
Werbad
|
New Level: Train gaparokI wanted to create a level that wasn't based on random barrels but that still would be a challenge. The resulting machine kinda looked like a train, so I made it into one. My record is at 4203 ticks, but I'm sure there is a faster solution.
|
|
|
|
|
Logged
|
|
|
|
|
|
|
|
|
jf
|
For Streams ( rilacog) I thought I had come up with a nice minimum-tick solution in fusutit. Then I realized that I was trying too hard: potygys (let it run for a few minutes). I think you mean when the machine gets stuck with F10 which happens if by chance two zeroes in a row are fed in at the bottom. I added the upper right block in my solution just to detect and handle that special case. For Streams ( rilacog): fobotedAnd for Change - Poor Version ( sugazuz): tabuson
|
|
|
|
« Last Edit: February 22, 2008, 04:11:10 AM by jf »
|
Logged
|
|
|
|
|
jnz
|
I think you mean when the machine gets stuck with F10 which happens if by chance two zeroes in a row are fed in at the bottom. I added the upper right block in my solution just to detect and handle that special case. Yes, that's what I meant. I was working on a special case for my machine when I realized that it wasn't going to get me what I wanted. I couldn't detect the change fast enough to stop the matcher from going red for a couple ticks. It wouldn't be truly solved until after we got stuck so why bother with the complicated machine when a very simple one could accomplish the same thing? Now if you could detect the change and switch modes without going red for even one tick ... For Train ( gaparok): beteguc. 1739 ticks.
|
|
|
|
|
Logged
|
|
|
|
|
jf
|
I think you mean when the machine gets stuck with F10 which happens if by chance two zeroes in a row are fed in at the bottom. I added the upper right block in my solution just to detect and handle that special case. Yes, that's what I meant. I was working on a special case for my machine when I realized that it wasn't going to get me what I wanted. I couldn't detect the change fast enough to stop the matcher from going red for a couple ticks. It wouldn't be truly solved until after we got stuck so why bother with the complicated machine when a very simple one could accomplish the same thing? Now if you could detect the change and switch modes without going red for even one tick ... For Train ( gaparok): beteguc. 1739 ticks. Here zigedam is a solution for Streams that detects the change fast enough to maintain the green matcher. The matcher may be red for the first couple of crates during initialization, depending on the start values, but once the matcher turns green it stays green even if the machine gets stuck on F10. This includes the case where the very first two ? crates happen to be zero. To force a stuck machine for testing, in the sandbox put a conveyor belt along the bottom from the right edge of the screen all the way to the bottom two ? crates (over the bricks) and load the right edge of the screen with tons of zero crates. They'll cruise along the conveyor and then jam the machine as they load in place of the ? copier output. Or, just change all three initial ? crates to zero crates, or even just the one over the copier, to vary the time before jamming.
|
|
|
|
« Last Edit: February 22, 2008, 08:58:16 AM by jf »
|
Logged
|
|
|
|
|
Werbad
|
I thought I had eliminated all getting-stuck cases for streams... Nice work solving it even with this error. For Train I noticed a design error that allows you to prevent the machine from locking the packer/unpacker sections... I will see if a can fix it... My solutions: Change ( hucyzir): jufobukRotation ( vagosuv): tuvudyrMessenger ( gegygif): meduxonEdit: Train v3 pesurudMade a second version of Streams. Fixed the double 0-case too. Streams '2' sixuruk
|
|
|
|
« Last Edit: February 22, 2008, 06:22:25 PM by Werbad »
|
Logged
|
|
|
|
|
jnz
|
Here zigedam is a solution for Streams that detects the change fast enough to maintain the green matcher. The matcher may be red for the first couple of crates during initialization, depending on the start values, but once the matcher turns green it stays green even if the machine gets stuck on F10. Ok, I couldn't let that go unanswered.  For Streams ( rilacog): xydonir. Solves in 41 ticks and stays solved.
|
|
|
|
|
Logged
|
|
|
|
|
jf
|
... For Train I noticed a design error that allows you to prevent the machine from locking the packer/unpacker sections... I will see if a can fix it... Edit: Train v3 pesurud... I already have a solution for Train, but it depends that design error for just one extra +. My best time is 2087 ticks so far, and I've just been trying to improve that. There is a solution, then, within the allotment of + and - ? If so, then I'll try that before looking at any solutions yet. EDIT> Oh, I see my mistake! I just noticed that the leftmost 0 is not sitting on a matcher. If you include the 0, then you need at least one more + or one more -.
|
|
|
|
« Last Edit: February 22, 2008, 10:06:23 PM by jf »
|
Logged
|
|
|
|
|
jf
|
For Train ( gaparok): huroroc in 1691 ticks.  I built a programmable machine which takes these opcodes: 1: add 1 7: no-op A: accept F: subtract 1 The paper tape along the top is executed as the two values loop, each in their own independent thread. Of the four possible sequences that solve using at most 15 + and 15 -, this is the shortest. The timing of the bottom track should work fine with Train v3 too.
|
|
|
|
« Last Edit: February 22, 2008, 11:49:56 PM by jf »
|
Logged
|
|
|
|
|
Werbad
|
Nice! My solution is also a kind of programmable machine, but a little more complex than yours...It is actually possible to snatch one more + or - by dropping the 2 crates at the same time...For Train v3 ( pesurud): lucibyb (3903 ticks) Is there nobody who has tried Doubles yet? Took a while for me to complete, but it is solvable... Edit: Merge notuberSurfing resugyd liladyn (fixed start of a dozer) Tiki Room gudefud
|
|
|
|
« Last Edit: February 23, 2008, 10:41:32 AM by Werbad »
|
Logged
|
|
|
|
|
|
Pages: [1] 2 3 ... 13
|
|
|
 |