On 6 Mar 1996, Joseph Strout wrote:
> > Sure. Get a bunch of TinkerToy kits and build a (manually controlled)
> > manipulator with which she can build other tinkertoy assemblies.
>.. but also felt free to design our own building blocks if
> necessary...
> way for the assembly head to get parts (perhaps a diffusion model...
> ...have to figure out how the head grips a block, then lets go of it when
How about liberal use of velcro on blocks of density 1 gram/cm^3 in a pool?
> We need to figure
> how the blocks connect; they have to go together easily enough that the
> assembler does not rip itself apart forcing blocks together.
> it's properly attached.
> Any ideas?
On a nanotech scale, things naturally want to stick together, which stretches the ananolgy with macroscopic replicators built of tinkertoys, which have to be forced together.
John Michelsen
[I think metaphors are being mixed here. If you're doing "machine-phase" assembly, all parts are in known positions at at times; if you're doing solution-phase mechanisms, it's not clear that robot-like manipulators are useful. --JoSH]