Oh hey, that’s me!

RE: https://mastodon.social/@rightardia/112441364366600074

me: “I wish people would stop calling individual blog posts ‘a blog.’”
*monkey’s paw curls*
https://xoxo.zone/@mathowie/112439941615950550

https://spectrum.ieee.org/ai-coding

Mixed feelings about this. Because on one hand, removing barriers to entry is fantastic. But on the other, I’m not sure whether this is an obstacle that’s actually necessarily gatekeeping or not.

In my CS class in high school, we were quizzed with programs with misleading syntax so that we’d actually develop the skills to read the actual code. I still don’t know if this is actually good pedagogy or not, but it is a skill I put to regular use when I’m coding.

Giving Aerospace a try since I’m still in the hunt for a decent WM on macOS, and this is looking pretty promising.

@zenspider Depends on what you mean - I pretty rarely reach for git bisection (except on one particular team/codebase - I think because it was a) needlessly complicated and b) important bugs were much later than when they were introduced).

Test case bisection I’ve only really used once in a codebase to find flaky tests - the rare codebase I’ve worked with that a) had a lot of tests but b) not enough discipline to find and fix the flaky tests.

But if you mean bisection via print statements? I do that shit all the time.

TIL the default Emacs calendar can
display the names of the months as they were used in Revolutionary France.

“The student left is the most reliably correct constituency in America. Over the past 60 years, it has passed every great moral test American foreign policy has forced upon the public”: Vietnam, apartheid South Africa, Iraq War, civil rights, etc. https://kottke.org/24/05/the-kids-are-right-and-alright

@ratkins This is why I think Microsoft is seemingly faring the best out of the major tech giants right now…

Today, 16 years ago, Debian published a security advisory announcing CVE-2008-0166, a severe bug in their OpenSSL package that effectively broke the random number generator and limited the key space to a few ten thousand keys. The vulnerability affected Debian+Ubuntu between 2006 and 2008. In 2007, an email signature system called DKIM was introduced. Is it possible that people configured DKIM in 2007, never changed their key, and are still vulnerable to CVE-2008-0166? https://16years.secvuln.info/

NYT: we can't voice our opinions without being shamed!

Everyone: which opinions?

NYT: ...

Everyone: WHICH OPINIONS?

CI / CD stands for “Continuous Integration / Continuous Disintegration”

I can’t believe Zoom has added so much cruft to the app that the freaking chat toolbar button has related to the overflow menu.

This article is about navigation in real life, but it would be interesting to see similar studies on navigating codebases. https://arstechnica.com/science/2024/04/why-do-some-people-always-get-lost/

@tahnok I’ve been wondering this too - maybe a combination of humans having worse color perception in low light and automatic image processing? 🤷

If you walk in San Francisco you've probably noticed more and more hostile "planters" blocking sidewalks. I'm part of a group pushing back on this, Gardens Not Barricades, and now we have a website/newsletter. Join us!

https://gardensnotbarricades.com/

#SFPol #HostileHorticulture

Yesterday I was a name-reader for my college's undergrad graduation, where we graduated ~900 students, aiming for a pace of 3 seconds/student. The students are asked to write their names "phonetically" below their name on the card we get handed. These are not linguistics students, and I really want to do research sometime on naive phonetic transcription of names. Somebody actually used "ough" to transcribe some vowel, who knows which one. Many used "aw" to transcribe /a/ (since they don't distinguish /a/ from /ɔ/ that makes sense). So you get "aw-naw" for "Ana." One student with a 5-syllable last name handed me his card and whispered "good luck!" He had actually done a nice transcription below it, probably using capital letters for stressed syllables, but I didn't have enough time to process what he meant about stress. I now have lots of thoughts about cross-language phonological processing in speech production, because the first choice in reading names is to figure out whether to use Spanish or English pronunciation, since we have a lot of students with Spanish names who would like them pronounced like a reasonable imitation of Spanish. A few of them just write in place of a transcription "Read in Spanish," which I find much more helpful than trying to figure out in real time from their transcription that they're trying to convey the Spanish pronunciation (ex.: "mah-dee-ah" probably means they want [maɾia] and not [məɹiə]). A student with the last name Nguyen gave me a phonetic transcription for their perfectly easy first name, but not for their last name. Thankfully, I do know how to pronounce Nguyen, or at least close. I also have thoughts about timing of phonological processing regardless of language--I'm sorry to so many students who wrote a perfectly reasonable transcription that I couldn't parse fast enough to pronounce their names right. Anyway, I find the name reading kind of fun. I'm hoping to do it again and to find a way to spend the 1 second or so before each student with their card in my hand instead of waiting for them to get to me so I can have that second for phonological/orthographic processing. #linguistics #phonetics

The United States is protecting is uncompetitive automakers addicted to expensive SUVs and trucks from facing a real challenge that would force them to innovate and could accelerate the EV transition by making cheaper models much more accessible.

https://www.ft.com/content/9b79b340-50e0-4813-8ed2-42a30e544e58

#china #ev #cars #politics

I thought I’d never see the day come. You can finally type `exit` without parenthesis to exit the Python repl.

Doesn’t work on Windows though.

#python #lmao

so, you’ve seen ™ and ™️ before. but like. why are there two. well, i have an explanation! the answer is: FE0F

first, unicode. unicode is a standard definition of a bunch of codepoints, where a codepoint is just a number with meaning. for example, unicode codepoint U+263A refers to ☺︎, or “White Smiling Face”, and U+1F431 refers to 🐱, or “Cat Face”

so, lets start by looking at the codepoints for ™. decoding it, it becomes the codepoint U+2122, referred to as “Trade Mark Sign”. this was added in unicode 1.1 in 1993, a decent time ago!

next, the codepoints for ™️. decoding it, we get two codepoints! U+2122 (™︎) and U+FE0F. wait. who is FE0F. why is he in my emoji

well, unicode isn’t as simple as a series of codepoints that refer to single characters. take a look at é̗ for example. this is three codepoints, U+0065 (Latin Small Letter E), U+0301 (Combining Acute Accent), and U+0317 (Combining Acute Accent Below). the first codepoint is simple enough, it’s just e. the next two, however, are combining codepoints. this means that they combine with the codepoint before them to modify it. U+0301 adds an acute accent above the previous codepoint, and U+0317 adds an acute accent below the previous codepoint. this example specifically isn’t very useful (i don’t know any language with a é̗ character beyond conlangs), but it becomes very useful for languages that use a lot of diacritics. imagine if we had to make a new set of characters for each set of possible diacritics! big waste of space, we shouldn’t have done that!

so, what is U+FE0F? well, it’s a special codepoint called “Variation Selector-16”. variation selectors are a reserved block of 16 unicode codepoints. only some have been defined, but among those currently in use are U+FE0E (VS15) and U+FE0F (VS16). from wikipedia: “VS15 and VS16 are reserved to request that a character should be displayed as text or as an emoji respectively.” so, what’s happening with ™️ is that it’s combining a U+2122 (™) and a U+FE0F (Variant Selector-16) to create an emoji version of ™. they’re the same character, just that one has been instructed to become an emoji!

also, for the interested, here’s the word “unicode” with a shit ton of combining characters: ù́̂̃̄̅̆̇̈̉n̖̗̘̙̐̑̒̓̔̕i̡̢̧̨̠̣̤̥̦̩c̴̵̶̷̸̰̱̲̳̹ò͇͈͉́͂̓̈́͆ͅd͓͔͕͖͙͐͑͒͗͘eͣͤͥͦͧͨͩ͢͠͡. what appears to be seven letters is actually 77 codepoints, taking up 147 bytes when encoded in utf-8. or 156 in utf-16. or 312 in utf-32. why does anyone use utf-16 if it’s longer? historical reasons :3

TL;DR: ™️ is ™︎ but instructed to be an emoji