I don't particularly doubt the waste of tape is somewhat necessary. The way TZe tape works is there's a double side sticky backing, a thin film with a coating of ink (selectively transfers, with the used tape re-rolled internally) when heated and pressed onto the sticky backing), and a clear coverlay. The problem is that because of the thickness of the cutter assembly (which is actually just a mechanized pair of scissors internally), there is about an inch of tape after the layers have been laminated, but before the tape is cut. The printer has to advance the tape to get it cut at the right place, but by the time it's gotten there, the tape now has a stub of white tape on it, which if not cut off, would get stuck to the next label.
"Chain printing" solves this problem by letting the printer delay the "advance and cut" operation until it has something to fill the otherwise wasted inch with.
This (along with the chain printing) only works when you know you have a lot of (similar) labels to print beforehand.
To me, it's also less about the cost of the wasted tape, and more about the extra litter it produces.
My idea is to add another endpoint to the API so that using Whisper or Siri, I can say, “Print Eggs and today’s date” or “Print Coffee in medium size.” I use something similar with an Alfred workflow, but voice commands would be nice in the kitchen.
I hate when someone talks about a project I’m interested in but doesn’t share any code. However, the entire thing is so ugly that I’ll never permit any other human being to see it. I’ll incorporate a Dead Man’s switch to delete everything or burn the Raspberry Pi, if needed.
Anyway, it’s all very simple once you realize that Fabric.js [2] handles 80% of the interface and functionalities on the frontend. Probably with Aider [3] and Claude Sonnet 3.5, you can make it during breakfast (or 7 years: I’m a dev, so I’m VERY bad at time estimations).
[1] https://gist.github.com/vsigler/98eafaf8cdf2374669e590328164...