mcovalt

If you own a Kobo e-reader, check out the Plato reader [0], an unofficial document reader of Kobo devices. I find it interesting that it's written by the same person that wrote the bspwm window manager.

It is one of my favorite pieces of software. The UI is beautiful and, to be honest, inspirational. It frequently reminds me that the best UI should not be noticed. Humbling.

[0] https://github.com/baskerville/plato

PostgreSQL also has built-in support for exact (as opposed to approximate) vector similarity search for some problems with the following constraints:

- you have smaller vectors (no more than 50D)

- Euclidean, taxicab, or Chebyshev distance is appropriate for your problem

Put a GiST index [0] on a cube column [1] and ORDER BY target_cube_value <-> cube_column.

[0] https://www.postgresql.org/docs/current/gist.html

[1] https://www.postgresql.org/docs/current/cube.html

Totally agree. I’d also say it paints yourself in a bad way since you’re showing you aren’t willing to work through your differences with Alice. No free lunch, but if you’re already committed to this approach the least you could do is be supportive of your colleagues even when you’re unwilling to work with them.

There's actually a right way to do this: pad it.

The blog post gives an example response to being paired with a coworker you dislike working with:

> "I'd prefer not to work with Alice again. There's a bit of a personality clash and I had trouble getting proper credit."

Don't just say this. Instead, say:

> "Alice is an incredibly bright person and a great developer. However, I'd prefer not to work with Alice again. There's a bit of a personality clash and I had trouble getting proper credit. I'm certain Alice will be a highly productive member of team in my absence."

If you don't actually think those things about Alice, take some time, meditate on it, and find some things you do really respect about them.

Is this the product that tree-sitter-graph [0] is being used for?

[0] https://github.com/tree-sitter/tree-sitter-graph

Note that the version of Firefox this article references was released on July 26, 2022.

The article specifically compares McMaster-Carr’s website to Amazon.com to demonstrate a superior UX.

Thanks for sharing this! I recently went through the process of generating vector tiles for the world at all zoom levels using OpenMapTiles. Seemed like a fun way to make my computer sweat for a while. The process was pretty straightforward, but thought I might as well share the resulting mbtiles file here if anyone wants it.

Link to torrent: https://kiwiziti.com/vector.mbtiles.torrent

The poor auto exposure and white balance touched on by this article really affected my webcam’s quality. A lamp in the background of my office caused my webcam to choose some setting that caused my face to be very red and blotchy.

I’m not being vain. People noticed it and it strained conversations. It’s harder to effectively communicate when your audience is uncomfortable by your appearance. Something to think about elsewhere in life.

But I digress. The Logitech software is horrible. I found CameraController [0] to adequately solve the problem. It allows me to adjust exposure and white balance. Now conversations feel more natural at work.

[0] https://github.com/Itaybre/CameraController

The intersection of graphs, FSMs, and linear algebra is really wild. I touched a very tiny bit on this topic a few weeks ago [0]. I highly recommend the book "Mathematics of Big Data" [1] if you're interested in this subject.

[0] https://news.ycombinator.com/item?id=31344146

[1] https://mitpress.mit.edu/books/mathematics-big-data

Congrats on making a decision sooner rather than later. I also dropped out of graduate school and remembered having some similar thoughts as you.

I remembered feeling a bit lonely, too. There's no one to talk to about your work since no one understands it. It's a bit obvious in hindsight. Research topics are at the fuzzy fringe of human knowledge. They exist just outside of others' research topics. Your job is to find something that makes the field even more complex. Or, even better, connect the dots that show it may not be so complex after all. In either case, no one knows what the hell you're doing. If they're in your little tiny field, they may have some vague idea about how it may relate to their own little universe of research. But most likely, when you come home and a loved one asks what you did today, you won't be able to have an answer and that can be frustrating. To compound the issue, everyone is familiar with what a PhD is, but few know how academia works. And there's some not-great aspects of it.

I've continued my research on my own time because it interests me. I no longer have looming deadlines and I don't have to check my bank account in the Wendy's drive-through line to see if my stipend will cover a chicken sandwich. I still value academia, though, and I know that my hobby-research will come nowhere close to matching the quality of research done with the resources provided by academic institutions.

I noticed this on https://hndex.org. So many searches for hair loss products. Like thousands... daily.

A weighted automata can be expressed as a sparse matrix and the matrix-vector product can be used to traverse the states. I think you could get a derivative out of that.

I dropped out of grad school where I focused on math/physics. I was too immature to effectively plan my research and write it up. Also, there's a lot to the institution of academia that I find very distasteful.

The courses were not really important, honestly. As opposed to undergrad, the classes were more provided as resources where you could get expert advice on the course material that applied to your research.

I'm still obsessed with the research, though. Linear algebra in particularly. I have had good success with keeping up with my little research interests by reading relevant new papers and books. I still plan on using what I've learned to advance the world in some way. Being able to work on it without looming deadlines and a shoestring stipend is freeing to say the least.

You really don't need to pursue a degree if you want to keep on learning.

> I was able to work around most of the cooling-related issues by mounting the laptop vertically...

Have you happened to measure the thermal difference vertically mounting your laptop? I also vertically mounted my home server laptop for the same reasons, but noticed the fans seemed much louder so I returned it to be horizontal like normal. It's as if they were designed so that the horizontal surface acts to attenuate their high-frequency sounds. It's still annoying, but less so. I'm wondering if going vertical again, but throttling the CPU as you've done is the way to go.

Very cool! Running the not-well-maintained https://hndex.org search engine (and other memory hungry linear algebra based services) was also my original motivation for tunneling to my home as opposed to hosting on a VPS.

Are you hosted via a residential ISP? It's my hunch that peering agreements favor routes of consumer -> data center -> consumer as opposed to consumer -> consumer. That's mainly why I tunnel. Has that been your experience?

This is fun. I'm more of a minimalist with my homelab setup. It's a laptop and an old NAS. I love it either way: running a homelab is a nonsensical and fun hobby in any case.

I feel like we live in a world in which it's either racks or cheap VPSs. In reality, at home, we have some serious CPU horsepower just spinning billions of empty cycles per second. Consumer hardware is insane.

I've handled 10's of thousands of unique visitors per minute and more than a couple front page of Reddit + Hacker News herds on this little laptop through a residential ISP.

Here's my setup (pic down at the bottom): https://kiwiziti.com/~matt/wireguard/

Sorry, confusing there. Other way around.

I see you use NixOS. Check out these NixOS modules for setting up what I'm talking about: https://gist.github.com/mcovalt/c1fc476385bd2b65513809c5bc68...