borse is a terminal program meant to practice reading braille, Morse code, and semaphore, which are common encodings for puzzle hunts. Also supports A1Z26 practice.

uvx borse

Or you can install from PyPI by using uv, pip, etc.

Configuration is stored in ~/.config/borse/config.toml by default, e.g.

progress_file = "~/.config/borse/progress.json"
words_per_game = 15
single_letter_probability = 0.3

Your daily progress is also automatically saved and displayed on the main menu.

Actually this is a note-to-self.

For Grade 1 (just the letters A-Z), the chart on Wikipedia is helpful! The trick is to memorize just the first 10 symbols for A-J, which only use the upper four dots. That's because K-T are the same as A-J with one extra dot, while UVXYZ are A-E with one extra dot.

In Morse code, the most frequent letters are shorter. So I think it's a lot easier to remember Morse code as a binary tree, since the common letters will all be towards the top. I found this picture on the Internet:

I think the dot looks like 0 and a dash looks like a (rotated) 1, so it makes sense to me that dots are in the left of the tree.

Then you can just memorize the letters in each row in order. Here are some terrible mnemonics I made up that worked for me for the first three rows (you're on your own for the last one):

• ET: Eastern Time, or a 1982 movie
• IANM: I Am Not Mad
• SURWDKGO: SuperUser ReWrote DynamicKernel in GO

(Also, surwdkgo is also the name of a Taiwanese CodeForces grandmaster.)

If you're learning Morse code for actual usage rather than puzzle hunt usage (meaning you plan to hear the sounds, not see them in a hunt spreadsheet) then in fact you need to practice with the sounds and not the visuals. It's simply too slow to try to remap sounds visually at 20+ WPM; you need to hear the letter directly from the audio. (In fact I think people don't even count the dits and dashes, they just hear the letter automatically.) So really in this case you should be using a real training rather than my silly puzzle hunt hobby project, e.g. https://lcwo.net/.

If you look at a semaphore chart, what you'll find is that there are some groups of adjacent letters that just differ in one hand rotating clockwise. For example, the letters from A-G are obtained by fixing one arm at 6 o'clock and rotating the other arm all the way from 7:30 to 4:30.

So in my head, I organize the letters in "blocks", where each block starts with two arms at a 45-degree angle, and then having the other arm rotate clockwise. The resulting blocks are then easier to remember:

• A block: ABCDEFG
• H block: HIKLMN (note J is missing)
• O block: OPQRS
• T block: TUY (note the additional Y)
• # block: #JV (this is the weird exceptions one)
• W block: WX
• Z block: Z

I don't know if A HOT #WZ means anything to you.

Set up by cloning the repository and running

uv sync
uv run prek install # pre-commit hooks

To manually run the linter and tests:

uv run prek --all-files # run linter
uv run prek --all-files --hook-stage pre-push

Dean Menzenes maintains a Rust port named borse-rs.

• Where does the name come from?

  From Braille mORse SEmaphore.

• Should "braille" be capitalized?

  No.

• Why would you spend time learning this?

  It could be a great conversation starter for a first date…