Theo Beers

I always get plenty because I’m the only one in my household who eats it. 4x ginger!

Unicode Collation for Dummies Learn how the Unicode Collation Algorithm solves the problem of sorting text beyond the basic Latin alphabet. A detailed overview with code examples in Zig.

I finally wrote a blog post to share some of what I learned from implementing the Unicode Collation Algorithm in two languages. Maybe there are people out there nerdy enough to be interested in stuff like this!

If I’d known I would be employed again at the end of the summer, I would have enjoyed it a lot more.

The one context in which I've ever used bubble sort for real. Does anyone recognize this function? I asked a few LLMs, and they all identified it at once.

Bonus question: how much better can this be done?

Steingass Persian-English Dictionary Improved digitized version of Francis Joseph Steingass’ “Comprehensive Persian-English Dictionary”

And here's a link to my Improved Steingass app itself, which I use on a daily basis. I think it's one of the best ways to look up classical Persian words online.

steingass.theobeers.com

For those who use Claude Desktop or similar MCP clients (a niche group, I know): I just wrote an MCP server to facilitate use of my API for Steingass' Persian-English dictionary. It works! A link to the GitHub repo with installation instructions:

github.com/theodore-s-b...

GitHub - theodore-s-beers/sangin: AI detection of Persian poetic meter AI detection of Persian poetic meter. Contribute to theodore-s-beers/sangin development by creating an account on GitHub.

GitHub repo with training data, scripts, etc.:

github.com/theodore-s-b...

I recently trained an ML model, fine-tuning XLM-RoBERTa and using data from Ganjoor, to predict the meter of a hemistich of classical Persian poetry. Still early days, but results are promising. Perhaps colleagues will find this interesting. Model on Hugging Face:

huggingface.co/katomyomachi...

Theodore as well 🙂

I went by خداداد in Afghanistan and عطاء الله in Egypt

Identitarian progressives bear significant responsibility for the destruction of American institutions that we are now living through. I will neither forget nor forgive how they drove the country toward this nightmarish situation.

crates.io: Rust Package Registry

On crates.io

GitHub - theodore-s-beers/gloss-word: A simple English dictionary lookup utility A simple English dictionary lookup utility. Contribute to theodore-s-beers/gloss-word development by creating an account on GitHub.

GitHub repo

Some of you may have noticed that Etymonline's website has been redesigned this year—multiple times, I think.

I try to keep up with changes to their markup so that my `gloss` CLI util still works (link in reply).

It also requires shortening of ناموخت to نامُخت for meter.

This line is identified by some as the opening line of Rūdakī’s mostly-not-extant versification of Kalīla and Dimna.

The most innovative feature is a collation tailoring that interleaves the Latin and Arabic scripts, so that alif sorts between A and B, bā’ sorts between B and C, etc.

Text Sorting Playground Try sorting text using different approaches—including mixing the Latin and Arabic scripts

If you want to test out how my UCA implementation works, you can use this little app I made earlier this year:

www.theobeers.com/allsorts/

GitHub - theodore-s-beers/feruca: Rust implementation of the Unicode Collation Algorithm Rust implementation of the Unicode Collation Algorithm - theodore-s-beers/feruca

The Rust library is here:

github.com/theodore-s-b...

feruca Python bindings for feruca, a Rust implementation of the Unicode Collation Algorithm

I recently published basic (really preliminary) Python bindings for my Rust implementation of the Unicode Collation Algorithm. My first PyPI package, at least for a project of my own.

pypi.org/project/feru...

Camino de Santiago? Probably yes

I have a little web app that gives examples of how Arabic numbers and their counted nouns work from 1 through 100, masculine and feminine, all three cases. This is of course only part of the story!

www.theobeers.com/arabic-numbe...

I have a little web app that gives examples of how Arabic numbers and their counted nouns work from 1 through 100, masculine and feminine, all three cases. This is of course only part of the story!

www.theobeers.com/arabic-numbe...

Depressing how DJT supporters are legit too ignorant to realize that what they’re doing, e.g. demanding the arrest of “activist judges,” is exactly what people descending into fascism have always done. Naïvely playing their role in a remake of an old movie they somehow haven’t seen.

I went to Princeton with Kratsios and was in the same eating club. He was one year ahead of me. Will Scharf, now White House Staff Secretary, was also Princeton '08, also in Charter Club (president of it, in fact). It all makes me sick. Kratsios was some kind of Thiel protégé.

I remember when I was a finalist for what would have been a life-changing professorship, I would think, Ok, it may or may not be me, but someone is going to be made very happy. It’s cool that someone gets to experience that. And in any case, I can find my own way. It helped me deal with it.

Möge Deutschland das tun, was Amerika ablehnt.

Frank pointed to the meaning of مات کردن as “to confound.” So that checkmate is not the death of the king, but his inability to make any further move. I’ve never taken the time to dig into this etymology myself.

F, split/moving back and forth, and always have been

What even is the lede? That Baker is leaving?

MAGA has ruined my family. We all still talk and get together, but it’s not like it used to be. We tiptoe around each other. I wonder if we’ll ever be able to put it behind us and feel normal again.