dan

i wrote about atproto and why it matters

Open Social — overreacted

The protocol is the API.

love that shit

my main “bluesky tip” is simply — like and follow posts and people you enjoy seeing in your feed, and “show less” what you don’t enjoy.

“training” Discover algorithm takes annoyingly long for now — but it does get more specific with time, and the team will be making it more responsive. data helps.

iceland

the video feed thing is a first pass that the team knocked out in a week. the algorithm is repetitive now — it will improve over time

i’m excited that third party devs can work with video feeds from day one. bluesky app can display other devs’ video feeds; other apps can display our video feed

in early 2000s microsoft ceo said open source was “cancer”. now open source is common sense

i wonder how long it’ll take for open social to become common sense

slow horses is better than i thought it would be. it’s something between The Americans and Misfits

how to fix 80% of bugs

1. find unnecessary accumulating state
2. derive it from existing stuff on the fly

so we better make this place work huh 😅

please keep liking this post. testing something

Reply to @danabra.mov

while i may do open source work in the future (whether react, bluesky, or other things), i’ll avoid any commitments on a foreseeable timeframe. i’m also not looking for a job. this year i expect my focus will primarily be on spending time with my family, touching grass, and being a lot more offline.

we’ve had many waves of people checking in on the site and bouncing. it always feels a little sad to see folks i know bounce but i know how sticky the twitter network effects are.

“build it and they will come” is not enough, but still — we aim to make the app better for every next wave, bit by bit

i know everyone’s tired of talking meta level but i’m really happy that i finally was able to cut over to posting just here. twitter felt pretty addictive even this spring and summer. but as our app got less wonky and we added features like quotes, it reached the good enough level for me personally

not malice, not incompetence, but a secret third thing

i can’t believe foundation has 86% on rotten tomatoes. the writing is like a bad episode of doctor who that thinks it’s game of thrones. but it is beautiful and there is something alluring about it anyway. so i will be watching the season three

Reply to @danabra.mov

i’m thankful to everyone in the community who are building things on top of atproto — you’ve been a real inspiration, and i hope this little experiment does work out. i wish all the best to my team, and i am beyond excited to not do anything remotely useful for a while. see yall on the other side :)

this thing in programming when you’re not sure where to properly fix the problem and then you think about it like, where is the first point that bad data enters the system. what it is the entry point for the confusion. this is where you mend it

found on wikipedia: the capture of russian explorers by samurai of the tokugawa shogunate in 1811

is there an rpg where you collect items, progressively level up, but after a certain point your character starts to mostly get weaker and eventually dies

poor craftsman blames his tools. good craftsman blames his ancestors

last severance episode confirms nobody fucks with the jesus

we deployed a fix a few days ago that should make Discover more responsive to “Show Less” for cases where it was previously ignoring it. hope it gets more relevant!

sneak peak of what just landed on main (should go out in one of the next releases)

▶

a few common Bluesky misconceptions i see in articles