@Lylat: I rejected 260 correctly, and it accepted it, so I'm not sure what happened with yours. Perhaps it's a bug. A natural power of 4 will consist of a single blue followed by an even number of reds.
A much simpler algorithm for Androids is to simply read one of each color on each pass through the string. I.e. Read a red first, dupe reds until you read a blue, then dupe to the end of the string and start over. It probably takes longer, but minimizes the number of parts.
Finally got a good solution to Ophanim. 97 parts, 35:23. Same basic technique as I used for Metatron - subtracted the strings bitwise, using yellow for borrow bits. Then it worked through the difference string in a separate routine to remove the yellows and determine if A>B or not.
@Emulous: The strings of colors accompany the robots. Sometimes you have to output a specific pattern of dots, while other times you just evaluate the string and accept or reject based on it. The criterion is written in white at the lower left. For level 1, the criterion is to accept every robot that shows up - just make a conveyor from the round start to the square exit at the bottom. For level 2, the criterion is to accept a robot whose string starts with a blue. So you need to design a process that will read the first color on the tape. If it's blue, send it on to the exit. If it's red, dump it to the side. The actual tapes don't show up until you hit the triangle icon to test your device. You have to plan you systems on a series of "if, then" branches - if the dot is blue, I do one thing, if it is red I do something else, the net result being to send the proper ones to the exit, while rejecting the others.
@Thecakeisnotalie: For level 3 (Robolamp), the piece has to go through three blue arrows to ensure you have at least three blue dots. Use three red/blue switches in series so that the blue goes to the next one and the reds just get burned off. The third blue arrow heads for the exit. Solution:?lvl=3&code=c12:5f3;p12:6f2;p12:8f4;c12:7f3;c12:9f2;p13:9f2;c13:8f3;c13:10f0;
@frostydraco: "Input" is what's coming at you on the paper tape - a series of red and blue dots that you have to process properly per the instructions. Click on the triangle icon to see what is going to come. The goal is to use the other tools to build a process that decides whether that string of dots fits the criterion specified. For example, if you're looking for all-blue strings, use a red-blue selector, which will read the first dot on the paper. Leave the red outlet empty to dump any reds, while use conveyors to recycle the part back above the selector to read the next dot. Use a conveyor to send it to the hole at the bottom from the gray middle arrow when you've reached the end of the tape. Hope that helps you get started.
Hah! Got Metatron in 119 parts, 19:29. Did it in two parts: first adding the strings bitwise, using yellow for carry bits, then worked through the sum string massaging out the carries until no yellows remained.
@ReynFox:
What on earth is going on with your solution to Orphanim? It clearly works and is very clever, but I can't fathom what it's doing or why.
BTW, I solved Judiciary by splitting the string in half, then sequentially comparing the first color in the 1st half to the 1st color in the second half, until I ran out of string. Not sophisticated or slick, but it worked.
Any hints on a general strategy for Ophanim (bonus level 2) would be greatly appreciated. I've contemplated trying to left-fill the shorter string with reds then comparing them bit-wise, or trying some sort of subtraction, but neither seems to be getting me anywhere. There's something obvious I'm not seeing.
Agree with many others. Pieces too difficult to move and place, tolerances for final shapes too tight - I have Level 10 completed, but it won't accept it due to small errors in exact placement of the boxes I can't get straightened out. Par times are ridiculously low, resulting in 0 points for nearly every level. Very frustrating game. It's not fun enough to justify the aggravation of moving and placing the pieces.
Unsatisfying that there's no "end". It just keeps going and going, until you finally just get bored and quit. Lots of potential here, not yet realized. I hope there is a verison II building on this concept.
slow + 3 splash in a corner is devastating. Upgrade to max. Add one air tower when needed. Between rounds upgrade air, splash, slow, and money. Start with as splash in a corner, add slow + 2 others. Easy.
This is awful. The things to click on are completely unobvious, the actions are illogical, the screen placement is too sensitive - and then it's not even the full version.
Completely dumb.
Good game, but it could have been great with more development. The leveling/skills/awards/heroes are largely pointless - they are completely unnecessary to completing the game. 5x maps would be needed to make the rest relevant. A great concept, inadequately executed, leads to disappointment over what might have been. I hope there will be a Protector 4 that will properly expand on the concepts begun here.
 "Manufactoria"){kind=icon}
 "Buzzle"){kind=icon}
 "Wyzau Circle TD"){kind=icon}
 "I Remain"){kind=icon}
 "Animal Cage Escape"){kind=icon}
 "Protector III"){kind=icon}
 "Escape Geek's Room"){kind=icon}
 "Farm Frenzy 2"){kind=icon}