Notices by pyrex (pyrex@dragon.style)

Icon (and handle) rotated!

It's this image by Skdaffle.

@inatus

Go ahead and draw me if you like!

(ref by Reilukah)

I saw a few people talking shit about Linux users.

I'm a Linux user! I'm also a Windows user and a Mac user. I try to make sure my software works on multiple platforms.

What I'm told is that I only use software that is blatantly broken in five ways, and I respond to bug reports by saying "fix it yourself."

This is surprising to me because my experience is that when I use proprietary software, it's full of bugs that go unfixed for years. The experience is often unfixably broken for disabled people, who aren't in the target market. But even for people who don't need accessibility features, I'm not sure the experience is any less broken.

I can understand feeling disempowered when someone asks you to write code or file a bug report. But I feel even more disempowered when literally no one in the world can help me. In many cases like Windows itself, the software is designed to hurt me, with irremovable ads and push notifications.

The actual business model is "no one can help you."

I think the reason that Linux software is perceived as buggy is because when you complain about bugs in Linux software, there is a significant chance they will be fixed. It is typically completely useless to complain about proprietary software and if you do so, you are probably doing so in a one-on-one, private medium. So no one will actually see your problem.

For that matter, it's not clear to me that Linux actually is harder to use than Windows. I think that when someone's bad at Linux, we say "Linux is hard," and when someone's bad at Windows, we say they're "bad at computers," which holds Linux to a double standard.

If you actually watch a new computer user learn Windows, you won't think it's easy.

It's also way harder for experienced users to do power user tasks on Windows because the environment will respond to basic operations like "launch Python" by trying to sell you on the desirability of the Windows store.

It might be tempting to say "novice users will not need to do that" but this is a major frustration in trying to support Windows users who have broken their system, often by using it as designed. And there's an even more fundamental problem in that nobody actually knows what Windows is supposed to do. There are relatively few tools designed to restore Windows to a known good state and Microsoft's preferred process for this will delete your data.

Even if it was easier to get started on Windows, I'm not sure it would be a good thing. Low tech literacy can be expensive: concretely, malware and scams rely on you not having the ability to distinguish them from legitimate content.

And Windows and iOS abet that. They actively train you not to understand:

- the distinction between privileged/non-privileged contexts
- the true source of messages you see from the computer
- the inherent danger of saying "yes" to things

I suspect the design of Windows itself is a primary reason for the success of ransomware.

(A footnote on accessibility: It's really common for me to see people reply to accessibility talk with comments that amount to "fuck the blind." To that I'll say: visual impairment has about a 2% prevalence rate. If you live to 50, you have about a 10% chance of visual disability, and at 75 that rises to 25%. So it's about as common as autism and rapidly becomes more common with age.)

Sometimes you just get an email with "undisclosed breach" vibes.

Stop Using Threads consumes a list of users and tells you which of those users interact with Threads. You can use it to remove Threads participants from your timeline.

You can find out the tool's opinion on you by entering your own name, or you can just Submit with default settings to find out about Gargron.

https://stop-using-threads.net/

Data practices, anti-abuse and opt-in/out are behind the link. If you didn't opt into search, you're not visible here either.

#threads #fediblock

A friend sent me this study on portrayals of mental illness in AAA video games.

To summarize:

- Identifiable portrayals of mental illness were, in all but one case, neutral or negative.
- Portrayals of mental health _outcomes_ were basically always negative.
- Mental health treatment was rarely portrayed at all and typically portrayed as ineffective.
- Schizophrenia-like illness was the most popular to portray, and schizophrenic people in games are usually violent.
- You are usually allowed to kill the mentally ill people.

That's pretty horrible!!!!!!!

Here's the study! (https://www.frontiersin.org/articles/10.3389/fpsyt.2022.967992/full)

@inthehands @arstechnica

unix_joe is just correct:

@thephd

I would love to hear someone else confirm that dtolnay did this.

I suspect someone may want to figleaf this by saying "it makes sense that the maker of syn would disagree with a technical approach that is very different from what syn did."

This is possible but would be surprising to me because none of the processes of technical criticism happened to you. They kind of happened around you, and the actual reasons for the criticism still have not been made public.

If this is true, it would imply to me that dtolnay just doesn't respect you as a technical contributor at all.

Which begs the question of why.

It's probably not related to your credentials or track record, you have those.

The Rust Project said it's not racism so it can't be racism.

I guess I'll land on "malevolent ghosts"?

@astatide @thephd

Still considering the possibility of a malicious changeling, or some kind of "ghoul elf."

My serious thought: yes, it's racism, and covering for the guy was more racism, and I look forward to the future where they continue to cover for the guy, which will continue to be racist.

(I've been the person who's had to say in meetings: "OK, it seems like you're not considering my proposal, but you're just changing the topic instead of offering a technical reason for it.")

@thephd

I suspect it would probably eliminate a lot of usages of dtolnay's other library, syn.

@thephd

WAIT WAIT WAIT.

Wasn't your work going to make the incredibly bulky existing proc macro infrastructure way less central to the process of building things like Serde to begin with?

It feels like you were working on a problem totally adjacent to the one dtolnay was trying to force adoption of his unpopular solution for.

EDIT: People are seeing this post. I am not a Rust Project member. Evaluate whitequark's opinion based on the level of credibility you assign to whitequark, but don't add any extra credibility to the claim just because I have repeated it.

hey @mastohost (cc @unascribed)

question that is not as hypothetical as you might think:

pursuant to the ongoing Mastodon Stampede problem (https://www.jwz.org/blog/2022/11/mastodon-stampede/)

have you considered running a mastohost-level instance of jort.link (or similar) and making your instances fetch their metadata from that?

i realize this would be a code change -- i don't think it would be a super big one though and i suspect the patch would be auto-applied. the infra-level burden of doing this seems worse, which is why i'm asking you before writing code