Dreamt of old brandy and house fires

[infryq]

Yesterday I was up early enough to put an hour into cleaning the needle and presser bars on the White. The needle bar is substantially easier to maneuver and is now satisfactory; the presser bar is still a bit stiff and scungy around the back. There’s more stuff I could unscrew on it that might free it up to give me more access, but I’m hesitant to do that as it’s not clear whether I’d be able to get the alignment right on reassembly.

For work, I finished the first pass of inference annotations and started fussing with how the information is presented. There’s the script, which is linear; ideally we could show just the script as a sort of summary of the scenario at hand. Then for each line of the script there are three kinds of annotations, some of which are extensive and/or have long lines. The inference annotations in particular can be expressed in natural language or as symbolic expressions. It makes sense to consider each kind of annotations in isolation or in various combinations. I played around with a spreadsheet layout, but the lines are too long to be wieldy. I almost want to make it a web app but that’s too heavy for this early in the project. I’ve sortof satisficed with a landscape document, an extra un-annotated copy of the script as an introduction to each scenario, and using tables for the natural language/symbol duality bits, which isn’t great but is probably good enough.

Over lunch and snack I found a couple video series from someone repairing an actual White Rotary 77 as well as a rotary from another brand but quite similar. He rambles quite a bit but that seems to be the price of admission for all these vintage sewing machine repair folks. I captured some good stills of his machine’s original knobs, and watched him demo removing and reinstalling the top tension mechanism, which definitely would have given me some trouble if I’d attempted it cold.

Dinner, crossword, DS9, no snack as it was late, sleep.