aezart

Microsoft is terrible about this kind of stuff. We have a big problem with MS Teams replacing tabs with nbsps in XML code snippet blocks. It breaks our pom files. We've also had similar issues with pasting excel tables into emails.

American tipping culture has its origins in the post-Civil-War south:

> Following the Civil War and the abolition of slavery, formerly enslaved Black workers were often relegated to service jobs (e.g., food service workers and railroad porters). However, instead of paying Black workers any wage at all, employers suggested that guests offer Black workers a small tip for their services. Thus, the use of tipping to pay a worker’s base wage, instead of as a bonus on top of employer-paid wages, became an increasingly common practice for service sector employment. In the early 20th century, these employers, who shared a common goal of keeping labor costs down and preventing worker organizing, formed the National Restaurant Association (NRA). Over the past century, the NRA has lobbied Congress to achieve these goals, first by excluding tipped occupations from minimum wage protections entirely, and later by establishing permanent subminimum wages for tipped workers (One Fair Wage 2021).

From https://www.epi.org/publication/rooted-racism-tipping/

The government staying out of it makes it worse. The companies have so much power over your life without any oversight.

A payment processor should be treated like a utility. They just let the money flow and skim profit off the top without caring who's at either end.

I think aside from any safety issues, another reason to prefer a deb or something over curl | bash is that it lets your package manager know what you're installing. It can warn you about unmet dependencies, it knows where all the individual components are installed, etc. When I see a deb I feel more confident that the software will play nicely with other stuff on the system.

I agree, and would personally extend that to all user interfaces that speak in first person. I don't like it when word's spell check says "we didn't find any errors". Feels creepy.

The first chicken Caesar salad I ever had was, I believe, at Metro Grill during the summer of 2006. I was not (and still am not) much of a salad fan, but that was the salad that made me say "maybe I can learn to like salad."

Interesting that sometimes a washed out color will look darker than a pure color, even though there's more light overall.

What's wrong with that is that a lot of languages don't call them hashtables. I don't think I've used the actual term since like 2016.

Arizona had AAA000 when I was a kid, then 000AAA, then AAA0000 issued sequentially, but then just before they got through all the plates starting with 'C' they switched to a new system that appears completely random to me. It's a shame, I really enjoyed tracking the progression and knowing whether someone got their plate before or after me.

Other people will occasionally verbalize the same kind of thoughts that I've had yet never spoken aloud. That's enough to make me accept that we're at least thinking in the same way.

I don't know what you're arguing against but it's certainly not the point I was making.

I didn't say "artists should be given money because they tried hard." I said that if you're profiting off someone's work, you should pay them.

Art is labor. Just because the end result is an "idea" with no scarcity doesn't mean that time, energy, blood, sweat, and tears didn't go into making it. I consider profiting off someone else's art without compensating them for that work to basically be wage theft.

How do you propose that artists make a living?

It's still entertaining, that's true. I like magic tricks.

The point is the analogy to LLMs. A lot of people are very optimistic about their capabilities, while other people who have "seen behind the curtain" are skeptical, and feel that the fundamental flaws are still there even if they're better-hidden.

> I have seen much AI output which is extraordinary it's funny how one serious fail can impact my point of view so dramatically.

I feel the same way. It's like discovering for the first time that magicians aren't doing "real" magic, just sleight of hand and psychological tricks. From that point on, it's impossible to be convinced that a future trick is real magic, no matter how impressive it seems. You know it's fake even if you don't know how it works.

Bulk mail should be banned, yeah.

Julia is referred to as "her" in the puzzle, so it's reasonable to assume she's a girl.

In humans, language is based on the state of the world, and anticipates future states of the world. In an LLM, language is just language.

If an LLM says "my favorite flavor of ice cream is cookie dough", it's just a statistically plausible thing to say.

If I say "my favorite flavor if ice cream is cookie dough", it's because I'm trying to deceive someone, because for some reason I don't want them to know my real favorite flavor is cake batter.

I don't know what the legal answer will be, but I believe it should be considered distribution. A model is basically a highly lossy and extremely compressed copy of its training data, available as a content-addressable database. To anthropomorphize, the model is trying to perfectly replicate its training set, its brain just isn't big enough to do so.

I tried 3.0 RC2 the other day and discovered that the new UI is much less space efficient. It imported my panel layout from 2.10 (which I spent a long time getting just right), and I had to completely rearrange it to get it to fit on a 1080px tall monitor. Very frustrating.

Also, it doesn't detect my drawing tablet.

I think I kind of get it. By the time someone actually gets to the point of filing an issue report, they are at the end of their rope. They have tried everything they can think of. They have googled and found no one else having the same problem, or fixes that don't work, or people saying "why would anyone need that feature". They feel like they're being gaslit, their time is being wasted, and that the developers are intentionally antagonizing them. And then the form to submit the issue has way too many fields and comes across as very adversarial.

That's certainly how I felt when trying to get my drawing tablet to work properly under Linux Mint, although in my case I skipped filing an issue and just gave up and went back to Windows.

It's the technology message board on 4chan, each board has a name like that. /a/ for anime, /v/ for video games, etc.

> (for example, children of obese people are more likely to be obese)

I would assume that this is related to gut biome and not genetic makeup.

Because it has no mass.

Too bad the rest of it hasn't been dogfooded nearly that much! The user experience on stuff like setting up animated sprites, tilemaps, etc. is still pretty uncomfortable. And all the UI stuff is very clearly designed for building desktop applications instead of, you know, games.

They can make it pretty hard to find the actual renewal price on domains in some cases.

If the domain is currently being sold at a "normal price", the search page will show the normal registration price and the renewal price. But if there's a promotional discount for the domain, the search page will show the promotional price and the normal registration price (to show how much of a savings you're getting), with no mention of the renewal price.

It's been a few years since I bought a domain from them so this could have changed since then.

> You may license other files that are entirely your own work and do not contain code from the software under any terms you choose.

This part confuses me a bit, as the distinction between storing all the code in the same file versus dividing it into separate files is fairly arbitrary. What happens if you have 3 files in your project, one of which contains licensed code, and then zip them together into a single archive for distribution?

The Godot game engine has its 2d graphics origin at the top-left of the screen, with the positive Y axis pointed down. Having clockwise trig functions is a natural consequence of that.

I think it's a fairly common setup for all 2D graphics software.

I don't understand how AI companies can claim that they're not aiming to output the training data when the loss function is "how well can model memorize the dataset?".