I’m still reading Sed Nur Fragmento by Trevor Steele. It is so good. As I mentioned last week, it really leans in to the Esperanto literature trope of hyper-focussing on language. For example, there is a scene where an indigenous person from the South Pacific is possessed by a demon. To highlight the fact that something super-natural is happening, Steele has the demon speak to the main character, Maklin, in Russian. That scene was so creepy, for reasons above and beyond language, that I got a bit nervous walking up to the bedroom with the lights off.
After finishing War and Peace on the Kobo, I essentially haven’t used it at all. I was thinking of various possible follow-up books to read after War and Peace. Immediately, my mind went to re-reading Anna Karenina. After a bit of pondering, I realized that I wanted more long, beautiful, reflective prose. I love Tolstoy’s omniscient, omnipresent, omnibenevolent narration. A sort of holistic approach to narration that doesn’t leave any stone unturned.
So, like, maybe I want to read In Search of Lost Time by Proust? It’s certainly a long read. I picked it up from the library to check it out and got through the first twenty pages, which are a long and somewhat obsessive reflection on going to sleep and waking up. I might stick with it, I’m not sure yet.
Jayne in Brief is so close to being finalized.
The svgs that latex2html produces look pretty good but there are slight issues.
For example, it chokes on some custom macros.
The line $@@F$ in Table of Notation shouldn’t look like that1.
It is spitting out $\LaTeX{}$ guts by accident.
That symbol only occurs in a handful of constructions,
so I built a little tool chain to make those svgs by hand.
(And add some dark mode support to the CSS. And update the icons. And a few other things.)
The whole build setup for the web version is quite fragile robust and well designed.
Follow Dave’s example, I ought to document how I did all this stuff.
That itself would be a little writing project.
This week, I’m writing up a multiple choice test for my course.
Typically, I don’t give multiple choice tests, so this is new for me.
Immediately, my mind goes to to making a
smoldering heap of scripts
software package
for assembling multiple choice tests.
Something to do versioning, answer key scrambling, question selection, etc.
It would be sweet to be able to say:
“Give a test and solution key with 30 questions, with this coverage, and scramble all the answers.”
I think that this is such a Linux user problem to have: “I’m doing something for the very first time, and I immediately want to generalize it and code it in a bunch of fancy ways.”
The braiding course continues to be a complete blast.
To quote from an e-mail to my students about these diagrams:
If research is the process of “getting confused and then getting unconfused”, we are currently at a high point of confusion. We have a method for converting braid words in to track plans which worked on the 2-strand coil braid and the standard 3-strand braid. Now that we’re trying it on the simplest loop braids (the 3-loop 4-ridge tubular), it is very messy and we’re very confused.
We braided two samples of 3-loop 4-ridge tubular starting with different hands. One sample had two loops on left hand and one on the right (briefly, “L2 R1”). The other sample has L1 R2. These samples looked very similar.
We wrote down the braid word for L2 R1 and setup our track plan machinery: https://pgadey.ca/office-camera/2025-10-21t10c28c41-04c00/
The induced permutation worked out the way we expected: (2 6 4)(1 3 5). It has two cycles and the track plan is supposed to have two tracks.
The very confusing part was this: when we followed the strands through the machinery, we got very complicated curves which we don’t really understand. See the bottom right here: https://pgadey.ca/office-camera/2025-10-21t10c59c51-04c00/ (On the left hand side of the board we did a weird overlap thing. Not sure where it leads.)
We called these curves the “strand orbits” or “doodles”.
It turned out that the curve for (2 6 4) was mirror congruent to the curve for (1 3 5). This is good because the tracks have different orientations (see Basic Instructions, Fig 1C.)
We don’t (yet) understand how to smooth out these curves and recover the track plan.
My guess is that here is a complicated “timing” issue with these curves. We need to carefully follow the strands along the curves to figure out what is going on. One of the things that we noticed was that some of the strands “pause” on the tracks. This is not evident in the diagrams and needs further study.
Another thing we noticed, which is hard to see in the blackboard shot:
- the passage 1—>5 has “length” 4 on the curve
- the passage 5—>3 has “length” 2 on the curve
- the passage 3—>1 has “length” 2 on the curve
E. read through his recent [journal] entries and commented that nearly all of them are about how he wished he wrote more. I laughed and said that if he read personal blogs, he’d find that most bloggers also write about how they wished they wrote more. Whether it’s online or offline, it’s the same, eh?
This error might, or might not, be fixed by the time this Week Note goes out. ↩︎
Published: Oct 24, 2025 @ 12:00.
Last Modified: Oct 24, 2025 @ 23:26.
Home / Now / Blog / Notes / Reading / Office Camera / Tags / Bookmarks / RSS Feeds / Top of Page
Thanks for reading! If you have any comments or questions about the content, please let me know. Anyone can contact me by email.