viggity

There are some fascinating theories on how writing media affects writing direction and glyph shape. A lot of southeast asian languages have a lot of curves/circles because they'd use a stylus on a banana leaf and the curves are less likely to run a split in the leaf.

I am suggesting that the major causative factor for life outcomes is genetics. Not "racist" property tax funding models. You can take the girl out of the trailer park, but you can't take the trailer park out of the girl.

The difference in outcomes isn't from funding/resources. https://statmodeling.stat.columbia.edu/2020/06/26/why-do-the...

I took 3 years of Japanese in HS (96-99). About 2 years ago I was doing a lot of work with genai and japanese typefaces. It was wild digging into how different the japanese web is. Back in like 2005, it was common to stylize english text by embedding it in an image and then applying drop shadows, etc. By 2022 everyone does the vast majority of that within CSS. Not in Japan though, I couldn't believe how much text content is still in image form.

I have always been kind of hyper aware of <everything>. My wife and I only dated for 6 months before we got married in sept 2019 (second marriage for both of us, we knew what we each wanted). But I definitely felt a bit awkward telling her about what I thought was coming with covid in late dec 2019. She was a bit suspect at first but polite and just went along with it because we were newlyweds and she loved me and gave me the benefit of the doubt. Holy shit, most everything went how I said it was going to at least through June 2020. AFAICT, she's still convinced I'm from the future.

Kinda like when someone said that instead of running for mayor, Bloomberg could have given everyone in the country $1M. A guest said it on NBC and Bryan Williams, nor the pundit had any intuition that it seemed grossly wrong. https://www.politifact.com/factchecks/2020/mar/06/msnbc/bad-...

Nice work! Out of curiosity, how are you deciding what the verse of the day is? i.e. is randomness weighted somehow? There are are ~31k verses in the bible, 23k in old testament, 8k in new testament. Is it 2.5 times more likely to be an old testament verse? Or, perhaps you're picking a testament at random, then a random book, then a random verse. Lots of ways to skin the proverbial cat.

Since the whole point of this is to limit LLM token consumption, it'd be interesting to see the results of prompts that use it.

I've seen a ton of people who just paste a CSV into a prompt and expect it to work well because they don't know any better, but the results are typically hot garbage. It's too repetitive, it can't memorize and/or process such a big chunk of data. Asking an LLM to use pandas to iteratively analyze some CSV works great, though.

These kinds of results seem all too common. Like, why? Are companies just too used to using their general business attorneys for it, and those attorneys are just ignorant? Hungry for extra billable hours?

I worked for a PoE lighting company for a couple of years. Yes, the ability to program your lights based on time of day, occupancy sensors etc was all nice for the end consumers. But the big advantage was that since it didn't use mains power, the owner of the building didn't have to hire a union electrician at eleventy gillion an hour to move some light panels around in a drop ceiling.

Does the second amendment apply to non-citizens?

I'm against the government jailing a visa holder for their speech, but revoking their visa is not jail.

celebrating an assassination of a political figure demonstrates that you are not fit to be in a position of power and employed by the government. Teachers, doctors, social workers, police/fire/ems... sorry, no, you just don't have the proper judgement. I'm against the government firing a janitor or a normal office worker over those types of posts.

Having a visa is a privilege, not a right.

In general, no. And to be clear, this is my general understanding and I don't have the time to look up the specifics, but I'm pretty sure lidar uses a "carrier wave" in much the same way old telephone modems did, or your IR remote for your TV. The carrier wave is how these sensors don't get confused. If you've ever been in traffic, and you get a stray, one off "YOURE TOO CLOSE TO SOMETHING" when you in fact aren't, it is because a carrier wave from another car happens to line up and you have a collision (pun intended).

two seconds of googling: https://www.digikey.com/en/maker/tutorials/2021/understandin...

1948 vs 2025 Jewish Population:

Algeria: 140,000 -> ~0

Morocco: 250,000 -> ~0

Yemen: 550,000 -> ~300

Iraq: 135,000 -> ~0

Lebanon: 20,000 -> ~40

Iran: 135,000 -> ~0

The book suggests two main reasons for this.

1. The reformation increased literacy/education in the populace to a greater extent in protestant areas, because you no longer needed clergy to talk to God, or understand the bible. Protestant countries have had better education for longer and it has a compounding effect.

2. The "Marriage and Family Program" (the "MFP")... protestant areas discouraged cousin marriage and levirate marriage much earlier than catholic countries, and it is still very common in the rest of the world. Consanguineous marriage is ludicrously prevalent in the middle east, it makes most of the rest of the world more tribal and you end up with compounding genetic defects. By making cousin marriage taboo, it encouraged children to move to a different town and made people less clannish.

The book examines this in multiple different ways, not just at the national level, but even within countries (provinces that are more catholic vs more protestant, and even within Germany, how far the city was from Wittenberg), as well as comparing third world countries that encountered catholic missionaries vs protestant missionaries.

My dad is a tax attorney/cpa, but is also a magician. My birthday parties were dope. It certainly wasn't a full time job, but he made enough money from corporate parties doing walk around magic to fund his habit, and magic is expensive (most videos teaching the trick are ~$100 or more). He has a magic library that if I had to guess is worth 6 figures.

I've tried getting into the craft many times, and if I had to boil down the essence of nearly every trick (especially for slight of hand, not necessarily stage magic)... Imagine the stupidest, dumbest, simplest explanation of the trick that you write off as "well certainly they're not doing _that_", and that is what is happening. The real art is that doing that simple thing is hard AF, they have to practice it a thousand times to make it look convincing. I never really got into it because it required too much dedication. And if anything, that makes it more special than something more cerebral.

I feel like this is a step in the right direction, but a lot of emotive text-to-speech models are only changing the duration and loudness of each word, the timing/pauses are better too.

I would love to have a model that can make sense of things like stressing particular syllables or phonemes to make a point.

I find that most companies find that to be a feature, not a bug. People are more likely to hit accept if they can only see a small chunk of the content.

Say what you will about Hitler, the man _did_ kill Hitler.

Anytime béziers are mentioned on HN, I feel compelled to share these absolutely incredible videos from Freya Holmér

https://www.youtube.com/watch?v=jvPPXbo87ds (73 minutes)

https://www.youtube.com/watch?v=aVwxzDHniEw (24 minutes)

I work in this industry, relatively new to it. There is a mind numbing amount of work that goes into font engineering.

it makes intellisense/autocomplete work a hell of a lot easier. LINQ in dotnet does the same thing.

the "serif" on the "n" is called an "exit stroke". You often find lots of glyphs that get an exit stroke (the "l" and the "i" come to mind, but it is most glyphs that have a single vertical stem, or on the right most vertical stem) when you get the italic version of the typeface.

This is mega levels of dope. Small UI/UX suggestion on your report... Take all of your table borders, and change the color from black to a light gray. The borders add value (communicate the organization of the data). But that data isn't as important as the text/numbers inside the table. "Visual weight should match information relevance"

I've posted many times about how my relationship with opioids (prescribed Rx, not street drugs) is improved when I'm good about taking my magnesium, which I started decades ago because it was supposed to help with atkins/keto. Many people have chimed in that what is likely happening is that Mg helps your ATP, and I'm probably just metabolizing it better. I've also seen other posts about Mg being an NMDA antagonist. If you're interested, those may be additional things you'd want to look into.

Suno could do this, even with a reference melody :)

Polite data viz recommendation: don't use black gridlines in your tables. Make them a light gray. The gridlines do provide information (the organization of the data), but the more important information is the values. I'd also right align the drive failures so you can scan/compare consistently.

If splines/beziers are of any interest to you, I could not recommend these videos from Freya Holmer enough.

https://www.youtube.com/watch?v=jvPPXbo87ds https://www.youtube.com/watch?v=aVwxzDHniEw