Profile for amin

Display name
Amin, minor deity of the legume realm
Username
@amin@polymaths.social
Role
admin

About amin

Fields

Web Search Engine
https://clew.se/
Timezone
UTC+8

Bio

Amin is short for Benjamin. He/him.

An opinionated, massively-parallel hobbyist and storyteller, I love my bicycle and fountain pens. Having just finished my degree in Communication from Texas A&M, I plan to spend 2026 as a full-time freelance writer and see where things go from there.

No matter how often I insist that I'm not an extraterrestrial being with dual citizenship from Titan and Venus on a deep cover mission to prevent the next Martian invasion, my friends never quite believe me.

Stats

Joined
Posts
17444
Followed by
1349
Following
335

Pinned posts

jump to recent

Recent posts

I finally finished this blog post, which was very difficult for me to write, but I think it says things I needed to say.

It's about how AI has personally harmed me and has been hurting my motivation to write for years.

If you're a writer, I do have some hope and vision for the future I've written here, and I'd like to share it for you. It's not all doom and gloom.

https://benjaminhollon.com/musings/stop-asking-writers-about-ai/

I look forward to hearing from people what they think. Hopefully this piece helps some people. (Boosts are very welcome.)

I've added a couple other implementations of my urlref concept to my blog post about it!

This is so cool! If you've got your own take or would like to chat about the concept/system with me, I would love to.

Here is my original blog post about urlref, with the new "other implementations" section: https://benjaminhollon.com/musings/urlref/

To those using Polymaths Pages:

@orbitalmartian had a problem where for some reason the file permissions of the files for his site were changed (not sure why) so the webserver couldn't read it. If you have the problem you should be able to fix it with this command:

chmod -R 750 /var/www/<your username>

If it does happen, also let me know so I can take a look at the commands you're using to upload to see if I can figure out what caused this.

I'm diving deep into the release notes for the new GoToSocial release candidate and there's some super exciting stuff, mostly which will help out my end as the server admin.

That said, here's one cool thing I found that doesn't show up in the "release highlights" list: https://codeberg.org/superseriousbusiness/gotosocial/pulls/4827

This will allow those of us who like custom CSS on our profiles to have styling for specific profile fields based on their names! That could be really cool, though I'm still thinking up exact use cases I'll try.

Now that I’m done traveling I’ll actually be able to respond in a timely way to people who were interested in trying Polymaths Pages, the free static hosting I’m now providing to members on my instance.

Just visit https://polymaths.page for instructions and send me (here or via email) an SSH public key and the names of any custom domains you want to use, and I’ll get you set up. I’m happy to help if there is any confusion or issues.

(Even if you don’t plan to significantly use it to host your main sites, you’re welcome to set it up to claim your <username>.polymaths.page subdomain and put something silly up there, like I did)

Good time as any to put out this notice, particularly to Polymaths.social folk—

I am shortly going to be traveling internationally across 13 timezones. Do not expect responses from me in the next… let's say 48 hours. I may check in during layovers but even if, say, the server is down, I may be too exhausted to deal with it.

So I'm finally actually making the "Polymaths Pages" free static site hosting I'd been wanting to provide for members here for ages. I'd had some grand visions, but have decided to just go much lower tech, which should also be more reliable.

If you're interested in being a guinea pig, lemme know. (Only polymaths.social members, please.)

Today's thought experiment for my web crawler development is about how to improve my adtech/tracker detection. Currently, I'm just blocking by hostname, which has a lot of false negatives.

What I want to do is allow importing a standard adblock list like EasyList… I've been thinking over implementation a bit (can't find any libraries that do quite what I want in Python), and I'm wondering if I could actually do this fully in the database. Postgres has pattern matching capabilities, and I did find a library that'll parse the lists for me, so surely I could get that working?

The question is if it would be performant, but I think it could be…

I'm close enough to actually finishing my lengthy rewrite-from-scratch of Clew's crawler and index structure that I can actually start having frontend ideas, so I've been doing some brainstorming…

Something I've been wanting to add for a long time is an "EXPLAIN" mode similar to Postgres's, where you'd add "&EXPLAIN" to the end of the URL and get a debug view of the results page that gives you breakdowns of how it did things like parse the query, rank results, apply filters, and so on.

I think I'll definitely add that; while beta testing (and maybe after) I'll put a link on the results page to enter that mode to make things easier for my testers.

(I'm not quite at the point of recruiting beta testers yet, but I promise I'll post when I am.)

My pet peeve: people saying "compliment" when they mean "complement". I know it's not a big deal, but aaaaaa it irks me. Possibly since trying to use "complement" implies aiming for sophisticated writing, as opposed to many other tpyos which could just be in ordinary conversation.