one step at a time…

Mr. Tambourine Man Bob Dylan Then take me disappearing through the smoke rings of my mind Down the foggy ruins of time Far past the frozen leaves The haunted, frightened trees Out to the windy beach Far from the twisted reach of crazy sorrow Yes, to dance beneath the diamond sky With one hand waving free

Мальчик Аквариум Сквозь можжевеловый ветер Сквозь пламя, чище которого нет В хрустальных сумерках Светом звёзд и светом ветвей Задыхаясь от нежности К этому небу и этой земле И наш сын говорит нам "Господи, приди и будь

four grid in particular is punishing with the super aggressive crop. so i think you pretty quickly learn it’s not a visual platform. compare to instagram or rednote.

fair enough. i just mean that bluesky has copied twitter ui verbatim and then never evolved visually anything because it’s too scary. and the parts it did evolve (feeds, labelers) make no sense to me design-wise. so i’m ready for more experimentation. and i personally prefer large images by default

i kinda hate the byrds version though. the last verse is the best and they just skipped it

sure maybe! i don’t explicitly say this in my posts but everything i’m posting is just my opinion. it’s fine for opinions to be different. to me the grid is there just because it was verbatim copied from twitter, not for some deeper reason. i want bsky app to starts visually differentiating itself

personally i think the cropped grid is a shit format. limited to four, doesn’t work for video or live photos, layout is kinda unpredictable

dorian electra — mr tambourine man

To promote experimentation I added support for format=json which makes it similar to docs.bsky.app/docs/api/app... endpoint:
- same query parameters - limit, cursor
- same response format: foryou.club/also-liked?f...

it’s not uncommon to rely on the grid as a meme format, kind of weird to break that retroactively. though in longer term carousel makes more sense

i haven't tried researching yet but i'd appreciate a clear model!

Bluesky is looking for a design research contractor

big plus if you know Bluesky well

who's the best design researcher you know? nominate them (or yourself)

like i get that they're stored wherever i store passwords, and they're somehow based on biometrics, but i don't have a mental model of whether they're device-specific or whether i can lose it, and what to do if i need to log in from another device

the funny thing about passkeys is i still don't know whether i understand them

Introducing a new way to explore Bluesky: foryou.club/also-liked

Enter a post URL to see what other posts people who liked it also like.

For example, enter the url of this art post: bsky.app/profile/alar...
and you will get a bunch more art posts: foryou.club/also-liked?p...

@atproto.com @danabra.mov I made a version of "Statusphere" from the tutorial, changed to use SolidJS and Convex instead of NextJS and SQL.
Tell me I'm cool and please let me know if there's somewhere I can share this for others who might find it useful
github.com/literalpie/a...

YOU ARE CLEARLY ON A JOURNEY THAT REQUIRES MORE HORSEPOWER THAN MATH

This is some extremely interesting nonsense

they blocked me

well, thanks for the conversation anyway, this was fun

i find them sometimes incredibly useful and sometimes maddeningly stupid. i don’t like that they’re managed by those companies and are centralized. i don’t know what the word “like” means to you here.

this is different from “sloppy” llm usage in the sense that the produced object is a mathematical proof. so it doesn’t matter if it’s sloppy or not. a proof either checks or it doesn’t. if it checks then it’s true, you don’t need to trust the llm. you’d still need to verify proof statement ofc

the thing about llms is that they don’t get tired and are good at fiddling at a problem when there’s good feedback from tooling. which proof checkers provide. so i think it’s possible that llms will make proof-based programming more mainstream. humans can’t be bothered to do that

well it’s not probably what you would expect.

the part that enables better guarantees and standards about software (to the extent it’s possible at all) is mathematical proofs. but programming with proofs is too difficult for humans to do so this hasn’t enjoyed much attention in mainstream coding

a microblogging website isn’t one of such systems by definition of what you want it to do

they require guardrails and context to wield them well but they can definitely produce good code.

basically computing is just too flexible. there are approaches that bring more structure into it where possible but each is a genuine innovation or a breakthrough and also takes a while to mature. it’s a developing field. ironically llms can unlock some types of hardening that were tricky before

but even without prerequisites, reading top down doesn’t tell you enough about program behavior. it’s too dependent on how state evolves over time. there’s even mathematical limits on what can be determined by reading the code (even if you had arbitrary amount of time like centuries)

understanding prerequisite systems’ internal behavior all the way down is not possible because the underlying systems go way too deep. it would be like several lifetimes of learning that get outdated every ten years

software is just too hard for humans to think about. the only way is to break it down into thousands of slices where people specialize in each slice. but the issue with that is that mistakes can compound and nobody sees the whole picture. how would you fix that?