Mastodon isn't perfect.
But the fact a social network exists that is completely free to use
has no venture capital investors
has no shareholders to answer to
has no growth targets
with a web interface with zero tracking cookies
and mobile apps with zero trackers at all
with ten thousand server administrators who donate their time for user safety
is - in my opinion - mindbogglingly cool, given the state of the world we live in. Not everything has to be shit. People make things better.
@juliobiason Fair enough! Aesthetic preferences are important.
@juliobiason What's the benefit in having the extra code?
A blogpost on @spritely about why we're launching our first ever supporter drive https://spritely.institute/news/spritely-launches-supporter-drive.html
We could really use your help! We're a small nonprofit, and everything helps no matter how much you give. We just crossed $15k but we have nearly $65k left to raise to meet our goal this campaign. Anything you can do to help is appreciated! https://spritely.institute/donate/
I woke up to see the @spritely campaign bar moved quite a bit overnight!
We're off to a really good start! But we've got a long way to go. If you're thinking of donating, we could really, really use your help!
@fleeky Coca cola released ai generated video ads. Adobe used ai art to advertise a tablet. This marker pen company tried to pass off ai generated images in their ads as though it was drawn with their pens. I see a lot of matrix spam with sketchy ai generated imagery.
@distributed So coool!
Social Reader is out!
@dripline This is a test reply
Tried generating APL but I think I can't trust any code that I can't audit :P
Might make more sense to practice my "write comments and prompt to convert to code" approach instead so I'm still the one "solving" and leave the syntax to the LLM.
At the same time, people are generally boring, admins are generally ethical, and snooping takes *work*. I've dug into the database to look at DMs once (?) in five years, and that was at the specific user's request. I've also been inside the surveillance capitalism sausage factory: corporate snooping is both far more invasive than most people realize, and also *ridiculously* lossy/noisy. We don't have the resources to fight a subpoena, but also no government has *bothered* to ask for our data.
@sarajw One way to check is to view the accessibility tree in your browser's devtools. It should show you the same structure a screen reader would see.
Maybe this year for #AdventOfCode I should use an LLM programming tool like Aider to solve the problems and see if it's able to do any of them.
tech ramble
One thing I like about the MST paper is the focus on sparse replication of CRDTs which I think is crucial to consider for any sort of distributed application that wants to do CRDTs for their data model with a large ish number of "documents" to modify. One thing MSTs don't help with is compaction of CRDT data to help deal with long histories. Unsure how one would best address that tho.
tech ramble
Bruh. Prolly Trees are just so much easier to use than Merkle Search Trees. Just the construction alone ends up being so much more simple.
Like, just reading the overview from Aaron Boodman is so much easier than the MST paper and construction/updating/reading is so much easier.
IMO ease + sequential reads are more important than "layers" being similar sizes and containing items.
MSTs are a step up from the CHAMP/HAMT craze though.
Occult Enby that's making local-first software with peer to peer protocols, mesh networks, and the web.
Exploring what a local-first cyberspace might look like in my spare time.