Programmatic api access

[lilmuckers]

Hi there!

I’ve set up one of these recently to use on my white board. I’ve added a few little roller bearings to it to lift it off the whiteboard so it doesn’t smear the ink - and other than that it’s worked brilliantly.

My use case is that I generate an SVG layout containing my days schedule on one side, and my todo list on the other with an optional diagram to put in the middle.

This has become an amazing part of my work day workflow. But I need to upload it the night before manually.

What is love is the ability to use my home server to generate the SVG, then auto upload it to the mural with fixed parameters at around 5am

And then return it to “parked” over on one side.

[nikivanov]

Hi! There are 4 areas of interest for this project:

• API access to Mural: this is exactly how the browser UI is interacting with Mural already! If you want to drive it headless (without UI), you'll need to look into setTopDistance, setPenDistance, uploadCommands and run. More on this later.
• SVG to commands generation: this happens all in your browser. Check out this test harness that renders an SVG to commands in Node, which you then upload using uploadCommands endpoint.
• Home position: currently, Mural goes home after drawing, and home is in the middle of the drawing area. If you want it to be, say, top-left, you can modify it here.
• Continuous drawing: Currently, when Mural starts drawing an image, it shuts down the web server, and then does a full firmware restart once it's done. That means that it will require re-homing on every run. I had functionality for "resuming" - that is, you can tell Mural the hanger distance and it will assume the belts are already in that position - when I was using slow stepper motors. Since then I've partially disabled this functionality, but it should be easy enough to restore it if need be.

Also, you're not the first one to mention this use case. I wonder - how do you keep the markers from drying out if Mural is always loaded with one?

[lilmuckers]

Oh amazing!

I hadn't looked too hard at the api calls mostly becuase i knew it was doing the svg->commands in the browser and hadn't really had the time/willpower/etc to go into that level of depth.

Now you've pointed it out i feel a bit dumb for not looking for myself :) - Because i'm a ridiculous person i might spend a bit of time trying to abstract it out fo the web interface as a generic import library for CLI.

I had some ideas (again, without looking at feasibility) of adding an eraser pad to the bot with another servo, so that it could also do a run to clear the board i'd imagine that could be another use for a multi-coloured pen option.

as for the drying out fo the marker up to now - i sorta wonder if you could put a pad on the "home" position and have the servo push the pen into that to keep it moist.