Selrak
TIGBaby
|
|
« Reply #260 on: May 24, 2010, 12:22:23 PM » |
|
Hello there PleasingFungus,
I’ve been playing your game, and I find it absolutely wonderful. That’s the kind of game I just adore. I am for from finishing, but on my way. However there seems to be a little flaw. It’s likely someone already told you so in a comment on Kongregate or something, but there are so many that it got drowned in the flood of other comments, and I don’t feel like digging through them all or post mine there and see it disappear as well.
It’s about the Robotanks level. There seem to be discrepancies. I do believe my machine, the second one I made, is perfect (all modesty set aside there.). However strings 31 and 32 (11111 and 10000) which are both strictly greater than 15 (i.e. greater than or equal to 16, since we’re only dealing with integers), are accepted but turn red when accepted ! The test would then fail, where it shouldn’t. DO you know what’s happening ? I do have an idea but maybe you already know …
Best to you no matter,
Selrak
|
|
|
Logged
|
|
|
|
PleasingFungus
|
|
« Reply #261 on: May 24, 2010, 03:57:43 PM » |
|
Ah... I see I was assuming that the old one being highlighted, it would replace rather than append. Not your fault! ahahaha... new issue.
I see you have swapped levels 21 and 22 (back, possibly). Oh well. So I copied my solutions in order to swap them over (copy to text file, change level number, paste back) so they are right and... no. Copying works (I can paste into a text file) but pasting does... nothing, well, what it does is pressing ctrl-v or right-click-paste closes the save/load window and does nothing else. I can't type into the save/load window either; any keystroke is causing it to be closed without any changes being made!
Yeah, I broke it. Got three bug reports this morning via Kongregate, plus this, plus probably more in Formspring. I am the worst at testing. New bug-release version is up; maybe it'll just work this time? (History seems against it.) ...add more 'rewards' (I finished Engineers early on, and was disappointed the military and judge paths didn't end with similar 'cutscenes')... Hm! I suppose the current branching-level-structure changes things a bit, but I always saw Engineers as the only actual ending. You're not the only one to see things that way, though - just saw at least one commenter on Reddit who thought the same thing. The trouble is that rewards would come either in the form of more artwork - which isn't really my specialty (could you tell?) - or in the form of more levels, for which I am largely out of both ideas and space. I'll let the idea simmer for a bit. (Glad you liked the game otherwise!) You should add another robot with 3 blues at the end of the tape.
I've been talking about this a lot on the Formspring. Basically, there's always going to be a way to make a machine that passes all the current tests without actually fulfilling the condition, but in a way such that just one more test will cause it to fail. This will always be true, for any reasonable number of tests. The random tests will fail this machine occasionally, and if I make them invisibly demonic, then they'll fail it pretty often. I can't offer any more than that! - Hello there PleasingFungus,
I’ve been playing your game, and I find it absolutely wonderful. That’s the kind of game I just adore. I am for from finishing, but on my way. However there seems to be a little flaw. It’s likely someone already told you so in a comment on Kongregate or something, but there are so many that it got drowned in the flood of other comments, and I don’t feel like digging through them all or post mine there and see it disappear as well.
It’s about the Robotanks level. There seem to be discrepancies. I do believe my machine, the second one I made, is perfect (all modesty set aside there.). However strings 31 and 32 (11111 and 10000) which are both strictly greater than 15 (i.e. greater than or equal to 16, since we’re only dealing with integers), are accepted but turn red when accepted ! The test would then fail, where it shouldn’t. DO you know what’s happening ? I do have an idea but maybe you already know …
Best to you no matter,
Selrak
It's the Input mode (which I'm assuming you were using to test) - it was broken. It's fixed - either now or in the next version, I forget. Thanks for writing in!
|
|
|
Logged
|
|
|
|
Jordan
Level 1
One time known as Flaming Pear.
|
|
« Reply #262 on: May 24, 2010, 05:02:51 PM » |
|
Nice idea, and I like the way the "robotic" look mirrors the game's content. Thumbs up.
|
|
|
Logged
|
|
|
|
man-man
|
|
« Reply #263 on: May 24, 2010, 05:10:50 PM » |
|
If the tests are, in part, randomised, do we lose the element of comparing test-run times? Surely that's only a consistent point of comparison if the tests are the same... unless I'm misunderstanding how the random tests happens.
|
|
|
Logged
|
|
|
|
Mask of Destiny
Level 0
|
|
« Reply #264 on: May 24, 2010, 06:50:50 PM » |
|
I found this game via Reddit and I quite enjoyed it. My biggest complaint would be that if you end up with a program that doesn't halt, you can't see what input lead to that condition without having been paying attention when that particular test started. For the fixed tests, you can always just go through again more slowly, but with the random tests you may not stumble upon it again. So I just finished metatron, it works but is really quite slow in its operation. Has anyone else finished with an elegant solution?
I don't know if it's more elegant, but it's a good bit faster (1:19) at the expense of having more parts (129): ?lvl=31&code=c6:2f2;c6:3f1;c6:4f1;c6:5f1;c6:6f1;c7:2f2;q7:5f4;c8:2f2;g8:3f2;q8:4f2;p8:5f4;q8:6f6;g8:7f0;c9:2f2;r9:3f1;r9:4f3;c9:5f0;b9:6f1;b10:1f3;p10:2f6;r10:3f1;g10:5f0;q11:2f6;c11:3f2;r11:4f2;q11:5f1;g12:2f3;g12:3f3;c12:4f3;p12:5f7;q13:2f0;c13:3f0;b13:4f0;q13:5f5;b14:1f3;p14:2f0;r14:3f1;g14:5f2;c15:2f0;b15:3f1;r15:4f3;c15:5f2;b15:6f1;g16:3f0;q16:4f4;p16:5f2;q16:6f0;g16:7f2;q17:5f2;q12:6f7;b7:7f0;c6:7f1;c7:4f1;c7:3f2;c17:4f1;c17:3f0;i14:7f4;c14:6f1;r18:7f3;c17:7f2;c11:6f0;c10:6f3;c10:7f0;c9:7f3;c9:8f0;c10:13f2;c11:13f2;c8:8f0;c13:10f2;c13:12f0;c13:13f1;c14:9f1;q14:10f2;p14:11f4;q14:12f6;r15:10f3;c15:11f0;b15:12f1;q16:10f0;g16:11f0;b17:9f3;p17:10f0;r17:11f1;c18:9f3;c18:10f0;c18:8f3;i14:8f4;c13:8f2;g15:8f2;b16:8f2;c17:8f2;c11:8f1;c11:7f2;c12:7f2;c13:7f2;c15:7f2;c13:11f1;c12:12f0;c7:8f3;p7:9f3;b6:9f2;r8:9f0;q7:10f0;q7:13f3;c8:13f2;c9:13f2;b6:12f2;p7:12f3;r8:12f0;c7:11f3;p10:11f1;r9:12f2;q9:11f7;c8:11f1;c8:10f2;c9:10f1;c9:9f2;c10:9f1;c10:8f2;q10:10f2;c11:12f0;c10:12f1;q11:11f2;b12:11f3;c11:10f1;c11:9f2;c12:9f1;c12:8f2;g14:13f0;
|
|
|
Logged
|
|
|
|
fleacircus
TIGBaby
|
|
« Reply #265 on: May 24, 2010, 11:59:48 PM » |
|
Wow this game took over my brain the last couple days. Good work!
|
|
|
Logged
|
|
|
|
Wyrm
|
|
« Reply #266 on: May 25, 2010, 02:51:15 AM » |
|
If the tests are, in part, randomised, do we lose the element of comparing test-run times? Surely that's only a consistent point of comparison if the tests are the same... unless I'm misunderstanding how the random tests happens.
I think it does the standard tests, times them for the total, and then does the random tests after. the final score board has the times for rnd tests greyed out, so I assume they're not included in the total. Also, I'm trying to compact my academics down as small as possible for later levels, can anyone see a way to earn some more space on the bottom right? (NB it works for arbitrary strings of red/blue/ green)
|
|
|
Logged
|
|
|
|
PleasingFungus
|
|
« Reply #267 on: May 25, 2010, 12:22:23 PM » |
|
My biggest complaint would be that if you end up with a program that doesn't halt, you can't see what input lead to that condition without having been paying attention when that particular test started. For the fixed tests, you can always just go through again more slowly, but with the random tests you may not stumble upon it again. Hm! That does sound annoying, but I'm not sure what I could do to fix that. If the tests are, in part, randomised, do we lose the element of comparing test-run times? Surely that's only a consistent point of comparison if the tests are the same... unless I'm misunderstanding how the random tests happens.
I think it does the standard tests, times them for the total, and then does the random tests after. the final score board has the times for rnd tests greyed out, so I assume they're not included in the total. Yep!
|
|
|
Logged
|
|
|
|
Uberubert
Level 0
|
|
« Reply #268 on: May 25, 2010, 03:28:16 PM » |
|
I've been hooked on this game for almost a week or so now. I love it! I even managed to get the nerdiest of my friends hooked on it. I would love to have more challenges though, but I guess that is what the custom levels are for.
Also, I'm unable to put crossing conveyer belts any longer. It adds quite a bit of a challenge to the more complex levels, not being able to cross the conveyers. This is both good and bad, since the game is challenging enough as it is. Is this a bug, or just an intended tweak?
|
|
|
Logged
|
|
|
|
Stereo
Level 0
|
|
« Reply #269 on: May 25, 2010, 06:09:28 PM » |
|
So I just finished metatron, it works but is really quite slow in its operation. Has anyone else finished with an elegant solution?
I don't know if it's more elegant, but it's a good bit faster (1:19) at the expense of having more parts (129): ?lvl=31&code=c6:2f2;c6:3f1;c6:4f1;c6:5f1;c6:6f1;c7:2f2;q7:5f4;c8:2f2;g8:3f2;q8:4f2;p8:5f4;q8:6f6;g8:7f0;c9:2f2;r9:3f1;r9:4f3;c9:5f0;b9:6f1;b10:1f3;p10:2f6;r10:3f1;g10:5f0;q11:2f6;c11:3f2;r11:4f2;q11:5f1;g12:2f3;g12:3f3;c12:4f3;p12:5f7;q13:2f0;c13:3f0;b13:4f0;q13:5f5;b14:1f3;p14:2f0;r14:3f1;g14:5f2;c15:2f0;b15:3f1;r15:4f3;c15:5f2;b15:6f1;g16:3f0;q16:4f4;p16:5f2;q16:6f0;g16:7f2;q17:5f2;q12:6f7;b7:7f0;c6:7f1;c7:4f1;c7:3f2;c17:4f1;c17:3f0;i14:7f4;c14:6f1;r18:7f3;c17:7f2;c11:6f0;c10:6f3;c10:7f0;c9:7f3;c9:8f0;c10:13f2;c11:13f2;c8:8f0;c13:10f2;c13:12f0;c13:13f1;c14:9f1;q14:10f2;p14:11f4;q14:12f6;r15:10f3;c15:11f0;b15:12f1;q16:10f0;g16:11f0;b17:9f3;p17:10f0;r17:11f1;c18:9f3;c18:10f0;c18:8f3;i14:8f4;c13:8f2;g15:8f2;b16:8f2;c17:8f2;c11:8f1;c11:7f2;c12:7f2;c13:7f2;c15:7f2;c13:11f1;c12:12f0;c7:8f3;p7:9f3;b6:9f2;r8:9f0;q7:10f0;q7:13f3;c8:13f2;c9:13f2;b6:12f2;p7:12f3;r8:12f0;c7:11f3;p10:11f1;r9:12f2;q9:11f7;c8:11f1;c8:10f2;c9:10f1;c9:9f2;c10:9f1;c10:8f2;q10:10f2;c11:12f0;c10:12f1;q11:11f2;b12:11f3;c11:10f1;c11:9f2;c12:9f1;c12:8f2;g14:13f0; Mine doesn't run quite as fast but it's a lot fewer parts. Right now the only optimizations to the design speedwise would be to mirror some sections so the test cases run through it faster it's very compact as is. ?lvl=31&code=c10:10f1;c10:11f3;r10:12f2;r11:8f2;q11:9f1;q11:10f0;q11:11f7;p11:12f7;g12:6f3;c12:7f3;c12:8f3;p12:9f7;q12:10f7;c12:11f3;b12:12f0;q11:13f3;b13:8f0;q13:9f5;c13:3f0;b13:4f2;g13:5f1;g13:6f2;c13:7f1;q14:3f1;p14:4f5;q14:5f1;c14:6f1;q14:7f7;r14:8f2;g14:9f2;c15:3f0;r15:4f0;y15:5f1;q15:6f1;p15:7f1;c15:8f1;c15:9f1;c16:3f0;p16:4f1;q16:5f1;c16:6f1;q16:7f3;b16:8f0;b17:4f0;y17:5f1;g17:6f0;c17:7f1;r7:4f2;g7:5f1;g7:6f2;c7:7f1;c8:3f2;p8:4f1;q8:5f1;c8:6f1;q8:7f7;r8:8f2;c9:3f2;b9:4f2;g9:5f1;q9:6f1;p9:7f1;c9:8f1;c9:9f1;q10:3f5;p10:4f5;q10:5f1;c10:6f1;q10:7f3;b10:8f0;g10:9f0;c11:3f2;r11:4f0;y11:5f1;g11:6f0;c11:7f1;c12:3f3;c12:4f3;c12:5f3;g12:2f0;g11:2f3; I like that it's heart shaped too. Since it's a little boggling to read this kind of thing, the metacode looks like - use green/yellow for carries and to mark the end of the numbers (a)g(b)g(g/y)(c)g, where a+b = c after code executes, leaving ggg(c)g - fork based on the last digit of (a), then (b), and (g/y), put it at the start of (c) as 00, 01, 10, 11 - copy out the rest of (c) - repeat until the first number is gone - loop through as if (a) is 0 if the second number's not done (this is the asymmetry at the bottom, a faster general solution could be reached for this but none of the test cases have different length inputs) - remove green/yellows and submit @ Uberubert you can still make bridges, it notes at the bottom of the screen that you need to hold 'Shift' while placing them. I guess maybe linking 'cross-conveyor belts' to the name 'bridge' isn't immediate. I like it better this way, I found myself having to erase wrong-direction conveyors more often than I want to do bridges. The per-test "best time" do not seem to be working. These are taken in the order I ran the versions of this machine. Manufactoria 1.13c! Chrome 4.1.249.1064 Flash Player 10 Can anyone beat this Robo-Spies? 15 seconds, 8 pieces http://pleasingfungus.com/?lvl=16&code=c12:7f3;c12:8f3;c12:9f3;c12:10f3;c12:6f3;p12:5f3;p13:5f7;p12:4f4;
|
|
« Last Edit: May 25, 2010, 07:06:14 PM by Stereo »
|
Logged
|
|
|
|
Mask of Destiny
Level 0
|
|
« Reply #270 on: May 25, 2010, 08:06:05 PM » |
|
Mine doesn't run quite as fast but it's a lot fewer parts. Right now the only optimizations to the design speedwise would be to mirror some sections so the test cases run through it faster it's very compact as is. ?lvl=31&code=c10:10f1;c10:11f3;r10:12f2;r11:8f2;q11:9f1;q11:10f0;q11:11f7;p11:12f7;g12:6f3;c12:7f3;c12:8f3;p12:9f7;q12:10f7;c12:11f3;b12:12f0;q11:13f3;b13:8f0;q13:9f5;c13:3f0;b13:4f2;g13:5f1;g13:6f2;c13:7f1;q14:3f1;p14:4f5;q14:5f1;c14:6f1;q14:7f7;r14:8f2;g14:9f2;c15:3f0;r15:4f0;y15:5f1;q15:6f1;p15:7f1;c15:8f1;c15:9f1;c16:3f0;p16:4f1;q16:5f1;c16:6f1;q16:7f3;b16:8f0;b17:4f0;y17:5f1;g17:6f0;c17:7f1;r7:4f2;g7:5f1;g7:6f2;c7:7f1;c8:3f2;p8:4f1;q8:5f1;c8:6f1;q8:7f7;r8:8f2;c9:3f2;b9:4f2;g9:5f1;q9:6f1;p9:7f1;c9:8f1;c9:9f1;q10:3f5;p10:4f5;q10:5f1;c10:6f1;q10:7f3;b10:8f0;g10:9f0;c11:3f2;r11:4f0;y11:5f1;g11:6f0;c11:7f1;c12:3f3;c12:4f3;c12:5f3;g12:2f0;g11:2f3; I like that it's heart shaped too. That is impressively compact! My first (mostly) successful attempt used a similar strategy (Yellow for carry), but it was way bigger than yours. Yours does seem to have a bug though. If A is longer than B it either rejects the input or goes into an infinite loop depending on the exact input given (I tested BBBGBB: 7+3 and BRBBGBB: 11+3). None of the fixed tests are in this form, but the random tests sometimes look like that. Can anyone beat this Robo-Spies? 15 seconds, 8 pieces http://pleasingfungus.com/?lvl=16&code=c12:7f3;c12:8f3;c12:9f3;c12:10f3;c12:6f3;p12:5f3;p13:5f7;p12:4f4; Not me. I doubt anyone can else can go smaller or faster either.
|
|
|
Logged
|
|
|
|
Stereo
Level 0
|
|
« Reply #271 on: May 25, 2010, 08:49:32 PM » |
|
That is impressively compact! My first (mostly) successful attempt used a similar strategy (Yellow for carry), but it was way bigger than yours.
Yours does seem to have a bug though. If A is longer than B it either rejects the input or goes into an infinite loop depending on the exact input given (I tested BBBGBB: 7+3 and BRBBGBB: 11+3). None of the fixed tests are in this form, but the random tests sometimes look like that.
I may have introduced a bug when attempting to optimize, it worked originally but I was messing with a little bit to strip leading 0s (the random code sometimes has them, heck so does the original blue green red - if those could be ignored it should be possible to skip tests) and may have left a mirrored GY switch. I'll check it out and see what's wrong. Edit: Here's one that works with both your examples as well as the regular tests. Metatron - 1:24 81 pieces http://pleasingfungus.com/?lvl=31&code=c10:10f1;c10:11f3;r10:12f2;r11:8f2;q11:9f1;q11:10f0;q11:11f7;p11:12f7;g12:6f3;c12:7f3;c12:8f3;p12:9f7;q12:10f7;c12:11f3;b12:12f0;q11:13f3;b13:8f0;q13:9f5;c13:3f0;b13:4f2;g13:5f1;c13:7f1;q14:3f1;p14:4f5;q14:5f1;q14:7f7;r14:8f2;g14:9f2;c15:3f0;r15:4f0;y15:5f1;q15:6f1;p15:7f1;c15:8f1;c15:9f1;c16:3f0;p16:4f1;q16:5f1;c16:6f1;q16:7f3;b16:8f0;b17:4f0;y17:5f1;g17:6f0;c17:7f1;r7:4f2;g7:5f1;c7:7f1;c8:3f2;p8:4f1;q8:5f1;q8:7f7;r8:8f2;c9:3f2;b9:4f2;g9:5f1;q9:6f1;p9:7f1;c9:8f1;c9:9f1;q10:3f5;p10:4f5;q10:5f1;c10:6f1;q10:7f3;b10:8f0;g10:9f0;c11:3f2;r11:4f0;y11:5f1;g11:6f0;c11:7f1;c12:3f3;c12:4f3;c12:5f3;g12:2f0;g11:2f3;g14:6f1;g8:6f1;c7:6f2;c13:6f2; A few pages back people were talking about Rocket Planes, here's the best I've got 18 pieces 1:54 http://pleasingfungus.com/?lvl=27&code=c10:6f1;b11:6f2;p12:6f3;q12:8f0;c12:9f3;c12:10f3;c13:6f3;b11:7f0;r9:7f2;p10:7f1;r10:5f2;q11:5f3;g12:5f3;c12:4f3;p13:7f3;r14:7f0;i12:7f6;r13:8f0; 16 pieces 2:00 http://pleasingfungus.com/?lvl=27&code=r12:4f3;r10:8f2;p11:8f7;c11:9f3;c11:10f3;b12:8f2;r13:6f0;q13:7f2;p13:8f5;r14:8f0;q11:11f3;b11:6f3;p11:7f6;c12:6f3;g12:7f0;c12:5f3; Introducing the extra red at the start makes it loop a little longer, but it also means the empty string is never handled so there doesn't need to be an exit from the first red-blue switch's "none". I'm not sure if there's a better way of handling that 'null' case.
|
|
« Last Edit: May 25, 2010, 09:28:44 PM by Stereo »
|
Logged
|
|
|
|
Derek
|
|
« Reply #272 on: May 25, 2010, 09:12:18 PM » |
|
Just started playing and this is pretty fun. I like the idea behind it a lot! But please, I really don't like playing puzzle games like this and having to use the keyboard. You should be able to click through the cutscenes with a mouse, and there should be a rotate button that you can click to rotate items. At least in Chrome the mousewheel causes the entire window to scroll. But yeah, having to use the keyboard for a few tasks when everything else is handled with the mouse drives me nuts. Mostly because I like to sit back in my chair when I'm thinking about what to do.
|
|
|
Logged
|
|
|
|
kenb215
TIGBaby
|
|
« Reply #273 on: May 25, 2010, 10:36:18 PM » |
|
I'm experiencing a glitch with the conveyor belts. If I put two conveyor belts on one tile the new one replaces the old one instead of becoming a bridge over it. This happens in both Internet Explorer and Firefox, and both at pleasingfungus and Kongregate. It was working correctly a couple days ago.
Edit: Didn't see the message on the bottom. Why did the behavior change?
|
|
« Last Edit: May 26, 2010, 07:50:17 AM by kenb215 »
|
Logged
|
|
|
|
Wyrm
|
|
« Reply #274 on: May 26, 2010, 02:03:58 AM » |
|
Can anyone beat this Robo-Spies? 15 seconds, 8 pieces http://pleasingfungus.com/?lvl=16&code=c12:7f3;c12:8f3;c12:9f3;c12:10f3;c12:6f3;p12:5f3;p13:5f7;p12:4f4; Mine is similar, but will take longer because mine works on RRRRRRRRRBRR (00004), while yours rejects instantly. Mine has the first component as a conveyer which means it strips off leading reds rather than instantly discarding.
|
|
|
Logged
|
|
|
|
man-man
|
|
« Reply #275 on: May 26, 2010, 04:12:32 AM » |
|
What's the best people have done with Robo-Tanks? It's not a difficult one, but I feel like I have more conveyor belts than ought to be necessary.
Any advance on 0:38 with 30 parts?
|
|
|
Logged
|
|
|
|
Doomsday One
Level 0
|
|
« Reply #276 on: May 26, 2010, 05:38:41 AM » |
|
What's the best people have done with Robo-Tanks? It's not a difficult one, but I feel like I have more conveyor belts than ought to be necessary.
Any advance on 0:38 with 30 parts?
0:38 with 28 parts
|
|
|
Logged
|
|
|
|
naath
Level 0
|
|
« Reply #277 on: May 26, 2010, 05:41:34 AM » |
|
Any advance on 0:38 with 30 parts?
0:36 with 30 parts :-) (I'm procrastinating trying to solve Metatron by working on making my other solutions smaller).
|
|
|
Logged
|
|
|
|
Wyrm
|
|
« Reply #278 on: May 26, 2010, 07:55:16 AM » |
|
What's the best people have done with Robo-Tanks? It's not a difficult one, but I feel like I have more conveyor belts than ought to be necessary.
Any advance on 0:38 with 30 parts?
0:38 with 28 parts 26 parts. I maintain time is pretty pointless; it's so very dependent on abusing the tests whereas a small solution is useful later on if you need to re-use it. Androids Robochildren - I meant robo-children! is a good one to play with. Try to beat 17 parts :D (and yes it works on any B/R input...)
|
|
« Last Edit: May 26, 2010, 10:25:19 AM by Wyrm »
|
Logged
|
|
|
|
JWP
Level 0
|
|
« Reply #279 on: May 26, 2010, 08:12:20 AM » |
|
What's the best people have done with Robo-Tanks? It's not a difficult one, but I feel like I have more conveyor belts than ought to be necessary.
Any advance on 0:38 with 30 parts?
0:38 with 28 parts 26 parts. I maintain time is pretty pointless; it's so very dependent on abusing the tests whereas a small solution is useful later on if you need to re-use it. Androids is a good one to play with. Try to beat 17 parts :D (and yes it works on any B/R input...) 25 parts
|
|
|
Logged
|
|
|
|
|