STOP DOING OFFICES
Edit: Apparent creator of this image, Pavel Samsonov: https://bsky.app/profile/spavel.bsky.social/post/3jvu3gud4hk22 /via @jsit
@rra sadly the client API is still a WIP but my colleague should be getting to it next week ish. I'll see if I can get a lil somethin done before then though.
technically it should be possible to do it witg curl, you just need to do the digest calculation and signing stuff with openssl or similar. Sadly I don't have ready examples of it cause we did it in js.
Here's how we do it so far: https://github.com/hyphacoop/social.distributed.press/blob/initial/src/server/apsystem.ts#L129
@reconbot I've been working on this for a while. https://blog.mauve.moe/posts/distributed-press-social-inbox
Tl;dr we figured out how to get the ap static bits working and made a minimal server just for the inbox portion
@rra @site Hmm are you self hosting your inbox instance? We have swagger docs generated at /v1/docs
Sadly swagger doesn't have the HTTP auth stuff built in so we don't have an out of the box UI yet.
We use the http signature spec for authentication so you shoukd sign yout request with the private key corresponding to your actor's public key: https://docs.joinmastodon.org/spec/security/#http
Mastodon should support Markdown formatting. This should actually be defined as a part of ActivityPub. Beyond bold and italics, it would be great to create links to make the content easier to read. https://www.markdownguide.org #Mastodon #ActivityPub
Still think it'd be useful to exted the #ActivityPub client-server spec with some sort of standard for "searching" through your messages.
IMO this will be crucial for making clients that can be interoperable between implementations (which I think is important for user agency!)
@pevohr Ah! We're putting in some last minute changes that might be breaking apis but I'm hoping to tag 1.0 within a week or two. Main thing right now is indexing replies so you can pull them into your post metadata.
It should be *mostly* stable :P
@mcollina Honestly I think your work back then is what set me on the path to making my own DBs. 🙇
@jonny Define hard and what sorts of indexes you want. IMO if you don't want a fancy query language it's not *tooo* difficult. You get a bunch of duplication of data though since you typically need to build like three indexes for querying either the subject or object or predicate.
i really love this presentation @mcollina did a while ago: https://nodejsconfit.levelgraph.io/
@jonny Real. Been thinking of making my own on top of leveldb since I kinda need one for distributed press stuff. :x also node.js's "level js" ecosystem means I could port it to browsers pretty easily.
@thomasreggi Yeah it's rough that the most "standard" js is for talking to external services rather than adding extra ones the way people do with Apache and the such. I guess one could plug node into a CGI script though :P
Personally I'm aiming to reduce the use of webservers in the first place and to rely more on static site publishing and client side rendering.
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.