tfw @aaronstevenwhite.io brings an analysis as sharp as a knife to your half-baked Saturday-morning thoughts: aaronstevenwhite.leaflet.pub/3miwsz2hdv22i 🤯
If we're going to own our data, let's actually own our data.
Which is to say: No, really, y'all, we're doing this. 💖🧠
Your rpg.actor has a new place to visit!
Right now you can head to the rpg.actor/theatre and watch videos of the @atmosphereconf.org direct from @stream.place while enjoying some FREE POPCORN, and chatting with others.
Enjoy exploring, and HAVE FUN!
More details here:
rpg.actor/news/the-the...
Wahh, this is amazing!
I've been working slowly on getting diarisation going, so we'll be able to hook up individual speakers to atproto ids. 👀
The talk timings will also be updated once I get a chance to re-run @iame.li's new talk VODs; my automated workflow took a bit longer than I was hoping.
I'm never gonna be an openai/anthropic/etc stan, but the fact that that post has 1k+ likes is so upsetting. Like, let's absolutely critique and even reject AI, that's fine, but do it with more finesse and poise than a gaggle of alt-right anti-vaxxers, please.
I don't disagree, but also see how that would be a very slippery slope from their perspective, since discover is "just a feed."
Which raises a question I don't know the answer to: is there a way to block yourself from a feed? (I would look, but I'm taking the cats for a walk 🐈🐈)
I guess all I'm arguing is that we need to separate the desire to express a preference from the question of how to implement that on the protocol and in our applications.
I would write it as a license/preference that says "don't show my posts in Attie-generated feeds"
To my more general point, this could be in an english document on the user's PDS, in a "licenses and restrictions" collection.
True, but the list of preferences is long, and that path leads the way to Twitter's Trust & Safety teams.
There's absolutely a short-term implementation question here, but I think we need to be careful about setting expectations about the shape of the network, especially in this moment.
Did you read my post yet? 😅 These are legal copyright licensing contracts; "own your data" implies this power (even if the license is non-sensical or unenforceable).
What we *do* about it is a separate question, and is going to take a whole lot more coordination power than we currently have.
There's something delightful about the fact that I can only see your posts here. Without going into the PDSes, I can't even see who you're arguing with. 😅👻
(Also, Bluesky should probably hide posts that @ someone who is blocked by or blocks the reader)
Omg. I really wish this was happening in June instead of July, or online. Can you get your lazyweb to remind me, too? Cc @adit.bsky.social
This follow-on explains my "don't wait for permission" stance a little; coordination on these sorts of platforms does matter, but I think we have a wide array of tools to do this incrementally vs. top-down: bsky.app/profile/blai...
Moreover, I'd claim an incremental approach is better!
You need to think about the recovery flow; passkeys are really inconsistent in that. For roundabout, we don't have a password recorded for any users. My approach is: email recovery is ground-truth, then layer opportunistic OAuth (esp Gmail), and then optional passkeys, esp for non-google accounts.
yeaaaahhhhhh!!!! 🧫🦠🪻🌾🌽🌲🌳🦉🦕🦎🐋🦧💃
I stopped short of trying to coin "Ecological Computing" in my talk, but I was *very* tempted. I love that it's happening here! cc @tessa.germnetwork.com
Really curious your thoughts on this: leaflet.pub/p/did:plc:3v...
Critically, the way I think about it isn't to just support outside forks – it's to allow major players (Bluesky) to make decisions without feeling the burden of being canonical or prescriptive.
I really need to open source Empathy Bureau. It's lexicon-agnostic pre-moderation with bounce-to-mod functionality, which is what we're doing (plus hacked up Ozone) for Roundabout.
oh, you, with practicalities like "writing things down" and "uploading records" 😂
also, have you written the 0008-72 lens yet? 👀
LOLOL. It took me a second to realize that you were replying to my post with your own post that you were [presumably] writing at the same time as me. So good.
giddy-up sparkle-farts!
Obviously, there are details I'm glossing over (300 chars isn't a lot), but @ngerakines.me already has the core of that implemented on lexicon.garden.
Totally. That's a big part of what's exciting about panproto - the lenses live on-network, and compose a graph, so there's effectively no implementation cost. You can just ask "I have a doc in format x, I need it in y" and panproto does it automatically.
I was going to go for a walk, but @rude1.blacksky.team nerd-sniped me into writing this, instead.
I'm not upset, though. It turns out collective governance questions are a great lens for grounding/scrupling how we might do on-protocol collective governance.
This doesn't get into the technical details (extremely lots here: panproto.dev), more the praxis framework for how I think we ought to approach questions like these: ionosphere.tv/talks/rj8Xv62
(the transcript should be mostly readable, working on some stuff to make it better)
a cartoon of a man with "sickos" written on his shirt, looking from the outside through a window to an unpictured interior, with a speech bubble that says "yes ... ha ha ha ... YES!"
Nope! That's still core. I wonder if @iame.li has a recording of the stream yesterday?
Did you check out the panproto.dev stuff yet? Coordination around lexicons is yesterday's game.
"This standards coordination meeting could have been a createRecord()"
We don't have to ask permission.
I'm very much a post-nerd atproto maximalist myself and would like more people of any level of technical understanding to get involved and mobilize around a locked-open system that even tens of millions in venture funding hasn't corrupted yet.
There's no way there's that many dots