Franken-Bots Connected bots go invisible

What's neat is how the boundaries get blurred with other such groupings.

Spoiler

 

I'm not sure if anyone has seen this phenomenon before but I noticed that if you connect R in this way either their legs glitch out or disappear the second the level starts. If the Robots are facing directions 90 degrees to eachother they disappear but if they're in the same direction or 180 degrees, they glitch out.

 

There is a level about this.

Spoiler

 

For designing my Valentine level I was using Franken-Bots without knowing that are called this way. Thanks to whoever invented this term which I like very much. The following video and level cards (based on that Valentine theme) I created as some kind of synopsis for this thread (notice also this other thread on the same topic).

• For directions please refer to this message.
• Trivial requirement: bots connected with irons to build a group and the group can (freely) rotate.
• When I speak in the spoilers below of rotations the rotational axis is parallel y axis. Degrees counted clockwise positiv and I like positive degrees more than negative degrees (you could read as well -90° instead of 270°).
• S bots = bots facing towards S; W bots, N bots, and E bots defined accordingly; F-Bots = Franken-Bots
• The arrows in the level point to the directions to which the corresponding bots initially look
  and the twin arrows point in addition towards S
  
• Any observation and conclusion below is only valid for the exact levels in the spoilers
  - I have got the impression that there is something more to the topic ...

Spoiler: S,W,N,E

• Connected S bots: group rotation 0°, bot facing direction preserved (bot rotation within group 0°), hovering for (short) period of time before gravity takes over, no extremity stretching
• Connected W bots, or N bots, or E bots: group rotation 90°, 180°, or 270°, resp., but bot facing direction preserved (bot rotation within group 0°), extremity stretching (legs, arms) due to group rotation: absolute feet and hands positions stay in play mode at same (or at approximately same) position as in design mode (the bots closest to the rotation axis, i.e. bots with smallest extremity stretching, might put their feet on some nearer ground which distorts the overall impression of the stretching effect somehow)

Spoiler: W & N

• Attach connected W bots to connected N bots: group rotation 180°, extremity stretching (legs, arms)
  
• W bots will become N bots at switch from design mode to play mode (bot rotation within group 270°), feet and hands positions rotated by 90° instead of 180° (extremity stretching not following group rotation nor same absolute position as before in design mode, notice additional offset)
  
• N bots keep their facing direction (bot rotation within group 180°), absolute feet and hands positions same as before in design mode (extremity stretching due to group rotation)

Spoiler: N & E

• Attach connected N bots to connected E bots: group rotation 270°, extremity stretching (legs, arms)
• N bots will become E bots (bot rotation within group 180°), feet and hand positions rotated by 90° instead of 270° (extremity stretching not following group rotation nor same position as before in design mode, notice additional offset)
  
• E bots keep their facing direction (bot rotation within group 90°), absolute feet and hand positions same as before in design mode (extremity stretching due to group rotation)

Spoiler: S & N

• Attach connected S bots to connected N bots: group rotation 180°, partial extremity stretching (legs, arms), easily to bring group in upright position (slight hoverring), instable movement when dragged
  
• S bots will become N bots (bot rotation within group 0°), no extremity stretching
• N bots keep their facing direction (bot rotation within group 180°), absolute feet and hand positions same as before (notice offset with regard to heart of S bots due to asymetric group)

Spoiler: W & E

• Attach connected W bots to connected E bots: group rotation 270°, extremity stretching (legs, arms), instable movement when dragged
  
• W bots will become E bots (bot rotation within group 270°), feet and hand positions rotated by 180° (extremity stretching not following group rotation nor same position as before in design mode)
• E bots keep their facing direction (bot rotation within group 90°), absolute feet and hand positions approximately same as before

Spoiler: S & W & N

• Attach connected S bots to connected W bots and to connected N bots: group rotation 180°, partial extremity stretching (legs, arms), instable movement when dragged
  
• S bots will become N bots (bot rotation within group 0°), no extremity stretching
• W bots will become N bots (bot rotation within group 270°), feet and hand positions rotated by 90° (extremity stretching not following group rotation nor same position as before in design mode, notice additional offset)
• N bots keep their facing direction (bot rotation within group 180°), absolute feet and hand positions approximately same as before (notice additional offset

Spoiler: W & N & E

• Attach connected W bots to connected N bots and to connected E bots: group rotation 270°, extremity stretching (legs, arms), instable movement when dragged
  
• W bots will become E bots (bot rotation within group 270°), feet and hand positions rotated by 180° (extremity stretching not following group rotation nor same position as before in design mode, notice additional offset)
• N bots will become E bots (bot rotation within group 180°), feet and hand positions approximately same as before (notice additional offset)
• E bots keep their facing direction (bot rotation within group 90°), absolute feet and hand positions (approximately) same as before

For the moment I would say:

• As far as I can see, the stretched Franken-Bots legs and arms
  do not interfere with bots nor with blocks nor with the ground.
• After switching to play mode all connected bots are facing same direction.
  This direction is descibed by group rotation + bot rotation within group
  = 360° (0° if using -90° instead of 270°).
• Group rotation and bot transformation:

• Connected S bots: 0° / S bots remain
• Connected W bots: 90° / W bots remain
  
• Connected N bots: 180° / N bots remain
• Connected W bots & N bots: 180° / all bots become N bots
• Connected S bots & N bots: 180° / all bots become N bots
• Connected S bots & W bots & N bots : 180° / all bots become N bots
• Connected E bots: 270° / E bots remain
• Connected N bots & E bots: 270° / all bots become E bots
• Connected W bots & E bots: 270° / all bots become E bots
• Connected W bots & N bots & E bots: 270° / all bots become E bots

• Group rotation in general: defined by dominant bot
• Hierarchy of bot transformation: E>N>(W|S)
• Group rotation in case all bots facing same direction in level builder:
  the number of right turns the bots are looking away from S
  equals the number of right turns the group is being rotated
  at switching from level builder to play mode. (Source)

​

I hope, there are not many errors in this overview.

References for further analysis:

• Edit mode order: S, W, N, E (Source)
• QR (decimal) block type for R bots = 26,
  QR (hexadecimal) orientation for block type 26: S = 00, W = 03, N = 02, E = 01 (Source)
• The way the app initializes bots for play mode (Source)
• Claustro path branching preference: E>W>S>N (Source)

 

Spoiler: Disappearing

Connected

• S bots & W bots
• E bots & S bots
• N bots & E bots & S bots
• E bots & S bots & W bots
• S bots & W bots & N bots & E bots

are disappearing, but their botic shadows remain.

The last combination of all four directions is the case for which this thread has been created.

 

I'm getting the impression that it is not only important which directions the connected bots are facing, it feels that the relative position of an individual bot within the group of connected bots has maybe the same or even greater importance. This way my 12 level cards above are just incomplete since they do not permute the bots position throughout the four hearts.

I'm focussing now on the last level card that connects all four bot directions. There are 4! = 24 possibilites to distribute 4 states to a 2x2 matrix, isn't? Not too sure, my math lessons are long way back.


I put the 24 combinations into 4 levels cards (below there is a video and the cards). From the 12 cards before I would have assumed that S&W&N&E connections would disappear. But in the new 2x2 example only 18 of 24 combinations disappear, 6 combinations don't. Obviously the order of the directions is having some impact.





The 2x2 geometry is only one of several possibilities to connect 4 fields. There are other side-by-side possibilities ... not exactly sure if I saw them all:

This sums up to 456 combination (in 2d plane) to create level cards for. Extending this approach to 3d would mean to deal maximal with 2088 combination :mad:, but due to symmetry this melts down to 1608 combinations - more or less

 

When the numbers are getting too big, it's better to go into the opposite direction. Reducing the approach from bot matrices to elementary bot vectors which only connects two bots. Three level cards have been created dedicated to the 3 basic directions south-north, west-east, and down-up. In the levels each two-botic and non-identical combination of S bot, W bot, N bot, and E bot, resp., occurs two times (with exchanged positions of the two individual bots; for identical bots, of course, there is only one vector). The video shows what happends to the connected bots in play mode (in the level builder the horizontal end of the curved rail points towards S in all cases). Disclaimer: since this all is only a fun hobby there might be undetected mistakes in the levels, but at the moment I can't see any, but well ... and in case there would be some corrections in the app with regard to the behaviour of connected bots, then all the analysis is only subject to some recycling process.

Spoiler: S - N

Spoiler: W - E

Spoiler: D - U

The following pictures indicate the directions the bots are facing in the level builder and also indicate the dominant bot (red) which defines the rotation of the connected bot group in play mode (in play mode the none dominant bot transforms to a bot that faces the same direction as the dominant bot in design mode). The group rotation is given in degrees as defined further above and disappearing combinations are indicated by X.



Where is the beef? I think, for these specific level cards:

• bots positioned South are more dominant than bots positioned North (S>N)
  
• bots positioned East are more dominant than bots positioned West (E>W)
• bots positioned above are more dominant than bots positioned below (U>D)

Remember: each botic pair is tested out in the levels two times. For example, W&E and E&W, with group rotation of 90° and 270°, resp., and the southmost, eastmost, or uppermost, bot is the dominant bot.

The orientation of the dominant bot is applied to the group of connected bots as a whole. Other (none-dominant) bots of the group have no influence on the group rotation? In most cases they don't but there are two exceptions:

• S bots cannot be transformed to E bots and vice versa
• S bots cannot be transformed to W bots and vice versa

The whole group of bots just disappear if transformations between S and E bots would happen, or the same for S and W bots. Transformation between S and N bots are possible. It might be that it is not possible to find the reason for these two exceptions (and to answer the original question for which this thread has been created) just by creating level cards that are testing out certain hypotheses. The phenomenological rules might be all we get out of this analysis. At least: the ordered group of orientations 01, 02, 03 (in additional to the base orientation 00) indicates that a transformation between 00 and the borders 01 and 03 of the ordered group is not supported whereas transformation between 00 and the element in the middle of the odered group is possible. Yes. So what? For further analysis, I would check to source code of the app. Would I really? No.

Above in this thread there is message #5 trying to identify a rule for this behavior and these exceptions. The level card "Connected R" of that message tests bot combinations S&E&N, E&N, E&E&W&W, E&S, W&E, S&N&E, and E&S (latter can be seen only in Mekostudio). The two groups combining S and N do also contain E and together with the S bot this additional E bot just makes the group disappear. Therefore it could not be seen back than that the S bots and N bots can transform into each other.

To explain the behavior of the 2x2 example in message #9 4 of the 5 atomic rules from the last 5 bullets would be sufficient (the 2x2 example just happened in the 2d plane, the U>D rule cannot be applied). In the 6 cases where the group was not disappearing always N bots were in the double dominant South-East position. In all other cases some not supported bot transformation between S bot and E bot, or between S bot and W bot happened.

The third level card for the D-U case shows something else:

• interestingly there are no rubberband limbs visible, the individual bots do rotate around themselves and from the observations before one could expect that the legs and arms of the rotated bots would remain on their position from the level builder, but in play mode legs and arms just appear normal, they do not break through the botic body to the opposite side, they just have been rotated together with the bot
• since the bot groups can freely rotate in the D-U level cards (in other cards the rotation is restricted to 90° steps) it can be seen that the initial rotation of the bot group does not introduce any actual rotation when starting the level card in play mode (use the draggables to play around with momentum in play mode), the initial rotation before starting the clock does not need to be slowed down, the initial rotation is just a change of position at t < 0 s, something different is when due to the initial rotation a block of the rotated group appears at a position where already before another block was located, then kinematics starts (t > 0 s)