Wha??

[Numsgil]

I'll be fixing the inability to revolve while hard tied in the next release or so.

[PurpleYouko]

I'll be fixing the inability to revolve while hard tied in the next release or so.

:blink:

Please explain!

You shouln't ever be able to revolve with respect to a hardened tie..... except by setting .fixang or one of the new .tiang() settings on the offending tie.

[Numsgil]

I'll be fixing the inability to revolve while hard tied in the next release or so.

:blink:

Please explain!

You shouln't ever be able to revolve with respect to a hardened tie..... except by setting .fixang or one of the new .tiang() settings on the offending tie.

I've been talking about this for a while.  Specifically I put it all together in the tie thread.

Bots in hardened ties will act as ball bearings.  Bots will only be able to change angles between two ties by applying nrg to move it.  That is, the angle between where the bot is looking and a single tie is meaningless for anything but the bot's own information.  That means the bot is free to rotate around while still manipulating ties, which is much better for reasons Ulc has clearly demonstrated.

The exact nature of the energy needed to move two ties I'm still researching, but it will probably be similar to the amount of energy muscles have to expend to move two bones in relation to each other.  You'll apply force X and drag and added mass from the ties will cause forces in opposition which can move you forward or to the side, etc. The force of these opposition forces depends on the viscosity and density (depending on laminar and turbulent flows) so is user definable, but forces will always be realistically based in physics.

Now you know what I've been doing for the last 4 days.

[PurpleYouko]

I've been talking about this for a while. Specifically I put it all together in the tie thread.

I must have been asleep then because I don't remember any of that.  :blink:

I can see your point about applying force between two ties but we also need to be able to fix an angle with respect to a single tie or else many structures lose their coherence. (note my reply to Ulc)

Maybe we should expand and refine the use of stifftie to allow various gradiations of fixed angle or rotatable ties since both systems have advantages.

[Numsgil]

The problem with that is that it doesn't take into account any of the forces.  That is, when you change the angle, what force is put on the ties and how much is just the bot spinning in place.  The only thing that would make sense is torque, but then your moment arm is so small you're having to spend alot of energy.

[PurpleYouko]

We already have torque.  :blink:

How do you think worms can swim?

Did you check out Carlo's Vermis P or my new swimmer test bot (in my last release)?

[Numsgil]

The bot gets a read back of his angle with respect to a tie, buth that angle itself doesn't effect anything in the simulation.  The robot is free to spin around without effecting the ties.

So you can orient yourself from a single tie, taking only the energy necessary to spin (which at this point is still free) or you can change the angle between two ties without effecting your own orientation.

It's the only model that makes sense (at least the only model I can think of that makes sense) from a physics point of view.

[Numsgil]

Yes, I know there's torque.  I meant that the torque of a bot turning a long tie just by the pivot would be very small.

I'm redoing all physics to make use of vectors.  Much cleaner (and nice informative variable names!)

[PurpleYouko]

If you allow all robots to spin freely on their ties then all the 2-bot MBs are going to be stuffed.
Inchworm relies on the stiffness and the fact that the ties spring back to the original orientation as do every other bot that I can think of.
This springy angle of ties is one of the few things that I really like about the way that ties work (and always have done) I would hate to lose the ability to use .fixang, tieang1 and the original tieang (which bends the tie temporarily then lets it spring back)

[PurpleYouko]

I meant that the torque of a bot turning a long tie just by the pivot would be very small.

Right. The individual robot should not realy impart much torque to the tie but the robot should still spring back to it's original orientation wrt the tie or else it becomes much harder to make small MBs

[Numsgil]

You just have to manually make sure you stay oriented to your ties.  Or we can even have a 'lock' variable that locks your orientation with respect to a tie.

Elasticity of joints isn't something I've explored yet.  I'm still researching this area.  But all multibot ideas I've had beyond simple animals, that is lattice connected organisms, needs to spin freely independant of tie angles.

Fish school, for instance, with bots connected to each other by ties so they don't have to watch each other to follow each other.

[Numsgil]

PY, can you think of a mechanical example of the tie structure as it stands?  Like my proposed system is a ball bearing.  Would help me look up physics stuff on it.

[Numsgil]

Ideally you could create a multibot that can feed at an angle to it's body's tie angle.  For movement you just have to rotate the desired movement of the multibot into your frame of reference.

Which again could use some cosines and sines.  Maybe there's another way.