2something

Emacs vs Ewindowss vs Elinuxs

I have a question for those who use screen readers.

Which of these pages is easier to read?

1) https://pandorastale.miraheze.org/wiki/Characters

2) https://pandorastale.wiki/Main/Characters.gmi

3) (If you have a gemini client installed) gemini://pandorastale.wiki/Main/Characters.gmi

#ScreenReader #Accessibility #a11y

When a calculus student is taking a test, and they have no idea how to do a particular problem, and the question isn't multiple choice, they have two options:
1) Leave the answer blank.
2) Write lots and lots of stuff in the hope the grader will see a tiny sliver of relevant information and award partial credit.

Option 1 means the student acknowledges that they don't know. That is a valuable skill in math. Option 2 is not at all useful for doing math, but is useful if you are a worker in capitalism and are trying to bullshit to a supervisor who doesn't understand anything. So of course many schools encourage students to take the worse option.

When students write bullshit answers on a test, their answers might contain some mathematically true statements. They may do some calculations correctly. But the calculations they do have no intrinsic meaning and no relevance to the question.

Now let's talk about big employers who are trying to "evaluate" their employees' "performance."

In many big corporations as well as many school systems, employees are evaluated using an opaque formula that takes in a bunch of data and spits out how "effective" the employee supposedly is. Employers can claim that because evaluations are done by a mathematical formula, they are "objective." If they give more promotions and raises to white cishet men, they'll say it's totally not discrimination: the objective evaluation formula objectively determined that white cishet men objectively did a better job.

The trouble is that there is no reason to believe any of the calculations that go into employee "evaluation" are in any way related to the thing they claim to be calculating. The people who write the employee evaluation formulae are doing exactly the same thing as the student trying to bullshit the answer to a math test: stringing together lots of irrelevant calculations they don't understand in the hopes of sounding impressive.

These are real decisions with major impacts on real people's lives, and they are being made with all the care and understanding that would earn a teenager a failing grade.

It's extra ironic when school districts use the same nonsensical formulae to evaluate teachers who understand why it's nonsense.

Now, I'm going to turn my attention to a group that doesn't get enough criticism: Animal Charity Evaluators. As the name suggests, Animal Charity Evaluators rates charities which they claim help animals. The people who run Animal Charity Evaluators are (or claim to be) utilitarians, so they believe you should donate money to the organizations which do the most good for the most animals per dollar.

In their review of charities, they use a Google spreadsheet containing a lot of calculations which they claim estimate how much good per dollar each charity does. But just like a student bsing a test question, they provide no justification for why the calculations they do have any relation to the global "good" function. Animal Charity Evaluators just ask us to take their word for it and give money to the organizations they picked.

Math teachers when I was a kid: Write down every step. Show your work!

Math teachers today: Use WebAssign, which only evaluates whether you have the right answer in the exact right format and doesn't even give you a space to write down your process.

I've taught coordinated classes where students have both paper assignments and WebAssign. I've watched students carefully write down every step when doing a problem on paper, only to see the exact same students struggle to do everything in their head when they try to answer a question on WebAssign.

Something about sitting in front of a laptop looking at the answer box leads them to not want to write down their steps. I don't know why: I'm not a psychologist. But I do know that the habits students are getting from WebAssign is really bad for doing math.

#FuckWebAssign #TeachingMath

If you hear a politician use the words "social media," you can find their actual meaning by substituting in the phrase "Ordinary people who are not celebrities communicating with other ordinary people who are not celebrities."

Politicians who rail against the alleged evils of "social media" are not concerned with opaque recommendation algorithms: they love those same recommendation algorithms when it's Spotify or CBS News using them.

Politicians aren't concerned about privacy: again, they love spyware when it's Spotify or CBS News or the police doing it.

Politicians do not view your self-hosted single-user blog any more favorably than Facebook: if anything they view you as more of a threat.

To the extent politicians dislike "big tech" at all, it's because "big tech" has historically permitted ordinary people who are not celebrities communicating with other ordinary people who are not celebrities.

There are no non-malicious reasons to ban ordinary children who aren't celebrities from speaking to other ordinary children who aren't celebrities. It's not a "well intentioned but misguided" policy. There are no "reasonable concerns." It's just pure evil.

Some advice for fediverse newcomers, particularly those coming here from corporate social media:

1) Follow and boost! Following is how you find posts you like, and boosting is how you let others find posts you like.

2) Alt text! Many fediverse users are blind. They cannot see images. Hence, if you make a post with an an image, it's good to include a text description of the image. This can be an "alt text," which screenreaders can read instead of the image, or it can just be text under the image. Some people are unwilling to boost posts that contain description-free images.

3) Content warnings! When you make a post about a heavy topic, you can choose to "hide" it behind a warning which mentions the kind of content. Anyone who sees the post can decide if they are in the right headspace to open it. For example, if you are going to talk about what a transphobic politician is doing, you can use the content warning transphobia, politics. Some users will only boost posts with heavy topics if they have content warnings, so using a CW can expand the reach your posts have.

If you're talking about a heavy topic, you should not try to hide it with "algo-speak." For example, if you're talking about suicide, do not say "su*cide* or "unalive myself:" type out the actual word you mean, and use a CW.

This also means if you see a post with a warning like "transphobia," it does not mean it's actually a transphobic post: it's likely made by a trans person critiquing transphobia. If you see actual transphobia, you should report it immediately. Speaking of which....

4) Reports! All moderation on the fediverse is done by actual living, breathing people! There's no AI scanning all your posts to decide who to ban or whose posts to hide. If you see a post containing bigotry or spam, press the report button (usually located near the reply or boost buttons) and write a report: the real people who run your instance will read the report and respond! I talk more about reports in the quoted post.

Also, if you receive a DM claiming to be a mod bot which tells you your account has been automatically disabled, it is a scam, and you should report it.

5) Money: Your instance is run by a single person or small group of people doing it for fun, in their free time, at their own expense. There is no company running your instance. There are no investors. There is no premium memership. There are no ads. You should keep in mind that the people running your instance are doing it for fun, and not to extract profits from you.

Your instance admins might accept donations, or they might not (it's their choice). If they do, and you can afford it, then it's good to donate to support your instance, but you will not unlock any additional features by donating. And if you can't donate, you should not feel bad about it.


#FediTips #Welcome

RE: https://transfem.social/notes/adhk91r22eiv0h80

Whenever I hear the word "omnipotent" I think it's a property of elements of a ring like idempotent or nilpotent.

I hate the phrase "brainchild" as a description of fictional media. It equates a work of fiction you create with a living, breathing human "child."

If you write a work of fiction and decide partway through you want to stop writing it, you can just stop. The fiction isn't a person with rights. It won't starve to death if you don't feed it.

Worse, I've noticed the phrase is used a lot by people who believe in "intellectual property:" These people believe a work of fiction is the "property" of the author's employer. If they describe a work of fiction as a "child," what does that say about how they view children?

What's the overlap between media publishers who believe "Trump says X" is news, and the people who think "AI says X" is meaningful information?

Did you know about the social contagion of Rapid Onset Job Dysphoria?

It's when someone who has always liked working long hours for sub-living wages suddenly starts identifying as wanting to be paid more or wanting weekends off.

I interviewed bosses that I found on rich-bosses-with-rojd-workers.com. They all talk about how their employees were completely happy with their jobs until suddenly they weren't. They also warn about a dangerous trend called a "strike," where a bunch of impressionable healthy workers all "come out" as wanting higher wages simultaneously.

#ROJD

Apparently folks are still confused about the purpose of sex education.

The first component of sex ed is teaching children
a) How to recognize if an adult is sexually abusing you, and
b) What to do if an adult is sexually abusing you.

Comprehensive sex ed includes other topics, including discussion of what changes your body will go through in puberty and how to have safe sex, but those often come later.

Sex education needs to start as early as possible, because if kids don't start sex ed until age X, then pedophiles know they can more easily get away with abusing kids who are younger than X.

The people most likely to sexually abuse kids are their parents, which is why we need schools, not parents, to provide sex education.

You likely can't teach a lesson about sexual abuse to a 1-day-old baby, but you absolutely can teach a 5-year-old.

When you hear policymakers say "kindergarten is too young to get sex education," what they are really saying is that they want to rape kindergartners and not get caught.


EDIT: In an earlier version of this post, I said that teaching kids how to recognize and report sexual abuse was the "most important" component of sex ed. After reading my own post again as well as reading y'all's excellent replies, I don't think I said that very well. Teaching kids how to recognize sexual abuse is an important part of sex ed, and it's the part that should be done the youngest, but the other components of sex ed (including what puberty entails and how to avoid stds/unplanned pregnancies) are also very important, and it doesn't make sense to single out one as the "most important."

Quick message to all fediverse newcomers:

Moderation in the fediverse is done by humans. There are no bots that shadowban you based on detecting certain keywords. This has some important consequences:

1) If you see a rule-breaking post, click the "report" button and write a report. The actual living, breathing creatures who run your instance will review the report.

The location of the report button varies depending on your client/app software: it's usually near "reply" or "boost."

2) The text you write in a report matters, because it will be read by an actual person. The mods are not just counting the number of reports.

3) If you're talking about a sensitive topic like suicide, you should not bleep it out like "sui-de". You won't be banned for typing the actual word out. You should instead use a content warning for your post to warn your readers. Some users who don't wish to see certain topics can voluntarily apply word filters to their own timelines, and bleeping words is rude because it bypasses their filters.

#FediTips