Bots and Simulations > Bot Tavern

Flying bot?

(1/2) > >>

Elite:
Is there any way of creating 'wings' using ties to make a flying bot in gravity sims.

I've seen bots in gravity sims form into huge tied masses which fall slower, but has any bot evolve or been created with the ability to fly.

What about something like this?


--- Code: ---o     o      o -- o -- o            o         o-o-o     o     o
  \   /                            /   \                  \   /
    o                            o       o                  o
--- End code ---


When a new 'flyer' bot is born it moves away from it's parent, repros, ties turns 180 degrees, repros and ties again. Then the middle bot then uses .tieang to 'flap', and .tielen to extend or retract, the wings.

Is it possible?

Questions:
If you have more than one tie, can you set the tie ID first and then .tieang or .setang or does .tieang and .setang affect all ties? What does .tienum control?

Numsgil:
Actually, I wanted to do something similar about 9 months ago, and found that the physics in 2.3X isn't stable and realistic enough to handle something like that.

It's actually a fairly complex physics problem.  I only really solved it (well, I think I solved it) a few months ago.  However, I haven't programmed it yet.  When I release the code I've been working on sometime in a month or two, it should be capable of handling that.

Endy:
I think the slower falling has something to do with total maxvel of the MB. It's some wierd stuff, bots falling slower if they have more mass.(DB is deffinitly not realistic here, a Galileo-bot would find that dropping a heavier mass falls slower :lol: )

There's also the oddball case where they'll fall much much faster, normally only when it's two bots connected and they form a horizontal line(I think, it doesn't happen everytime so I'm not sure here.)

Those oddball tie and mass physics get in the way right now. I know what you're saying though, I've been wondering if a human designer could manage to improve those oddball flyers the simulation makes. :D

PurpleYouko:
Here is the reason for MBs falling slower (as far as I can tell)

When two or more bots are tied together, all independent movements by any one of them due to any forces (their own movement or gravity) are mitigated and reduced by the effects of the ties.

Example: Bot A and Bot B are tied together.
When the program sequentially reaches Bot A to apply accelerations due to gravity, first the acceleration is applied then (since Bot B hasn't moved yet) it is found to stretch or shrink the tie, thus invoking the springiness to which applies an opposit acceleration on Bot A to bring it back toward the stable threshold of the tie. It also applies a small force to Bot B at this point so that they both tend to move toward the point where the tie is neither stretched or shrunk.

The result is that the actual movement of Bot A is much less than it should be.

Next it is Bot B's turn and it does the same thing. At this point though, as Bot B moves, it again changes the equilibrium of the tie and applies forces to both bots to bring them back to the point of rest. This means that both Bots experience a double damping effect.

I spent months trying to figure this out a year or so ago but as you can all see, it still doesn't work right.  :(

Numsgil:
I'm pretty sure I've solved how to do MB physics so that you don't have weird problems, but I haven't coded any of it yet.  If nothing else it should be less jerky than at present, and so hopefully more controllable.

Navigation

[0] Message Index

[#] Next page