Having all sorts of long-term ideas go out the window. Up until now I’ve been using a solenoid lock to stop a ratchet wheel in one direction and another ratchet wheel and arm to keep the assembly only turning in one direction. I had a limit switch that would flip on a turn.

This worked in principle but felt bad. It flipped the switch in a tiny turn. It turned forever in one direction. It didn’t feel like a doorknob.

I switched over to an actual doorknob inner assembly, that allows turning a quarter turn in either direction with a spring return. Then I take the limit switch and set it to reverse- send a press when released, so that the regular door bolt that is usually extended will send a button press when it turns and retracts.

This introduces problems for the solenoid lock. You can’t ratchet it in one direction because the turned knob needs to return. So the solenoid lock can stop one direction turn when locked but not turning in the opposite direction. This means that I’d need two solenoids to stop the knob from turning on lock and allow turn and return on unlock. Since the solenoid uses power when it is unlocked, and as it does generates heat, and the solenoid would usually be unlocked, it would be hot and powerhungry to use two solenoids. It would also take up space inside the plinth.

You could probably work out some complex ratchet-arm/servo solution but it wouldn’t have the satisfying “thunk* of the solenoid.

I think I am going to have to abandon the electronic lock and just represent locked doors with sound and graphics in unity.

SIGH.

Another cool physical thing that I spent a lot, A LOT of time figuring out, gone.

SIIIIIGHHHHHHHHH.

In other news, unity program is coming along, and I have an idea of what the copy-protection book/screen interaction will be. Going to prototype that next.

I took a first pass at scripts to get Grotto room info from the API and make a LaTeX file for printing a book with each room as a page, but my attempts at selenium scripts to scrape images of the canvas of each room (which I broke down and asked Chat-GPT for help with) is a bust, as it’s saving blank images, even though I can right-click ‘save image as’ on the canvases in chrome with no problem.

Book scripts are private but here, request access if you are an advisor/peer etc