I realized that I never wrote my #introductions, so here it goes.
I'm a Python programmer working on open source. After hours I do whatever is my current obsession, so far I have gone through: #wiki, #drawing, #typography, #roguelike #gamedev, #pygame, miniature #painting, #tinwhistle and #ocarina music, #arduino, walking #robots, #rcheli, #electronics, #microbit, #micropython, #circuitpython, homebrew game handhelds.
So yeah, turns out I made a tiny little mistake in the design, swapping two rather important pins: SPI MISO and SCK. So I fixed the design, fixed the firmware and fixed the prototype by swapping those pins back.
Oh, and I also have the Feather version assembled and flashed, but I still need to do something about cable management on that one. That's because I didn't want to cut the servo cables to size, and instead plugged them with their regular plugs. That will probably need to change at some point, unless I find some way to tuck them out of view somewhere.
This one can take any FeatherWing as a face, so I can use any of the LED matrix or display FeatherWings.
Not much time over the weekend, but at least I got the servos connected and their positions roughly tuned. Fine-tuning will be in software.
I also have a 4-digit 8-segment display that will fit as the face nicely.
I like to think that there is a corporate meeting room just behind that large dark octagonal window. #brutalism
Don't treat an inventor of something as an expert about it. They *invented* it, which means that it didn't exist before, so they really didn't have much chance of becoming experts. Experts come much later, it's those for whom the thing is a mundane tool to be mastered, not a novel gadget to patent.
The inventor might have spent time and effort researching it, and later, how it can be used. But they will remain at a disadvantage compared to those who started with the thing already existing and researched.
All legs and the electronics are now assembled. CircuitPython is flashed. I still need to connect all the servos.
It would also include all the other parameters in that scope, so that you could make the default values depend on them.
The PCBs arrived today, so I assembled one and compiled a UF2 bootloader and CircuitPython firmware for it. It works as expected.
I'm really happy that Elecrow used black silkscreen with the yellow boards. Most fabs use white, which doesn't look half as cool, in my opinion. And I can do the industrial-looking diagonal warning strips with that!
The parts are here, and they are perfect. So far I assembled the body and one leg to see how it all fits together. The holes for the servo horns in the femurs are too big (I was really afraid they would be too small), but the horns are held in place by screws anyways, so it's fine. I'm soldering up the PCBs now — there is one minimal, and one compatible with Adafruit FeatherWings. If I find enough servos, I will make both versions.
This will take a bit more time. Elecrow has "shipped" the PCBs and laser-cut parts last Monday, but they spent 8 days in a "shipment information received" state (which basically means nothing has been shipped yet, just the shipping label has been generated). I hope that now that it moved, it will be here in the promised 3 days...
The (PC) filesystems are horrible databases. Why are we still using them?
They can only be queried by the primary key. There is no query language, and any other searches must be done with external tools that literally iterate over the whole database. There is no indexing, no relations. Transactions are per-operation. No collision resolution. The entries themselves have no structure, just blobs of raw data. Extremely crude access controls. It's literally the level of technology of late '60s.
The social network of the future: No ads, no corporate surveillance, ethical design, and decentralization! Own your data with Mastodon!