>>729
Yes brilliant thinking, GreerTech.
On that topic: "Universal Robowaifu Software" is a rather-lofty goal, as one might imagine. Easy enough to gloss over details if you're an Asimov, or Heinlein, but realworld engineering always requires realworld solutions. Since physical systems always come down to physical items, physical design & testing is always involved. This is where the proverbial "Rubber meets the road" (and since we're dealing with electronics here -- at the atomic scale).
>tl;dr
We'll need significant funding, and a team of engineers working with several complex testbenches to manifest this dream solution properly. Till then, we'll need to create a series of specific solutions; keeping hardware-abstractions & generality, etc., always in mind. As the 'Lanky Dutchman' is fond of saying:
>"Create concrete solutions first, then abstract."
<--->
As an example of this approach: M$, et al, have already solved this concept in large part via their software drivers abstractions for general hardware & computing devices. So yeah, this idea's been around for quite a while.
We aim to make this process much simpler (ie, no custom driver software needed) for robowaifu creators -- by having them simply fill in a cleartext config-textfile template(s) that adapts his specific robowaifu's electronic (SBCs, MCUs, comms interfaces [I2C, CAN, etc.], actuation driver electronics, sensors, etc.) & physical (limb sizes, limb masses, torso length & mass, actuator performance aspects, etc.) characteristics over to the generalized (C3) robowaifu 'brain' software. That way, our engineering teams can stay focused on developing out the brain's command/control/communications aspects, and individual adopters can focus on crafting their cool waifu designs. Put the two together and its realworld abstractions win-win! :^)
<--->
But it certainly won't hurt to begin creating ad-copy & prop from that general platform! I'll be giving this exact point some thought soon. Thanks again, cheers Anon. :^)
Last edited by chobitsu