Joe Ruelle

Its a gravity problem, whatever it is it has the escape the gravity of bsky dot app. That’s Jupiter level gravity at this point.

You’re not signing the posts. You’re signing off on state changes to the tree. 🌳

But this would pretty much declaw did:webvh in a bluesky context no? You update your ICANN domain and you're still more or less back to square one (or so it seems)

I like the Touch Bar. There, I said it.

That I'd say is the key point.

Payments. For Bluesky it can only be payments in the end. But that has to be architected in at every level, a real challenge given all the other stuff on Bluesky's plate these days.

If you look at the history of startups that end up having to spend a lot of money every month but that aren't earning any real money, even after some years, then the odds are against them.

Any startup can still pull it off though.

ah hah, thanks

What benefit does having atproto in the mix here bring, I'm a little confused?

What do you think is the best proof of control, one for which you'd have no way to deny that you had that control for a given time period?

Perhaps a cautionary tale of what can happen when the community clients of a decentralised social protocol can't escape the gravity of that protocol's flagship client.

www.binance.com/en-IN/square...

@germnetwork.com Any plans to have Germ interop with other MLS apps? As far as I know there's MIMI for HTTPS, and Nostr has its own relay-based MLS interop standard (there are now 3 Nostr MLS apps looking to interop with each via that route). And maybe there are other MLS inteop standards out there?

I'm also thinking it's looking gloomy. This is a useful breakdown of what seems to have gone wrong that I just found (have to Chrome translate it from Chinese):

www.theblockbeats.info/news/57856

Penny for your thoughts on Farcaster? It's often seen as the odd one out. App is pretty slick though.

Great paper! Maybe a footnote that Nostr, thought not first envisioned with aggregators, has many in practice, the main example being Primal’s “caching service”, Ditto.pub the same, plus paid aggregator relays like docs.nostr.wine/filter/readme, and dev offerings like relay.nostr.band/index.html.

I've no idea, I subscribe to this bot and it told me about it. bsky.app/profile/cshc...

paper comparing decentralised protocols

@laurenshof.online Seems up your alley arxiv.org/pdf/2505.229...

Looks like Farcaster's flagship Warpcast client just rebranded as, well, Farcaster. What's going on I wonder, does this mean no more separation of protocol and client branding over there?

I think for those for whom it's a smaller, less functional Twitter, if you add a one-click move your repo anywhere button, all it becomes is a smaller, less functional Twitter with another button.

Got it, thanks, feels like I've hit atproto timestamp bedrock now.

To backtrack a second, the signed commit for the root itself, so change of state of the entire repo, is decent enough cryptographic proof of the state of that entire repo at a given point in time yes?

Gotcha. From the root what can you Sherlock Holmes out vis a vis the time an individual record was created? Since it's a change of the state of the entire repo (could be multiple/empty commits) I'm guessing best is a window? Or maybe write multiple records in succession for a kind of triangulation?

I guess question is how to verifiably "unpin" a keypair at a certain point in time should it become compromised, so that only signatures during the "pinned" period can be cryptographically linked to the DID? I haven't gone deep but I suspect timestamping at the PDS record level also has its quirks.

Ah hah, very clear thanks!

Sorry DID docs, can updates be timestamped in any way? With the goal of being able to determine when key x was added and when it was removed.

Researching how timestamps work in that context is on my to-do list. Or maybe that’s something you’re on top of?

I kind of see atproto as being like a yacht and nostr as being like a jetski on the back. There are a lot of places atproto just can’t go but that nostr can zip around to easily. And vice versa, you wouldn't want to live on a jetski. It's all about the integration layer, and also tribal barriers.

If you forget about the bluesky and nostr branding and just look at the underlying tech then it's all just open-source lego blocks.

If for signing stuff in the wild you could have a PDS record claiming ownership of a Nostr keypair and then use that keypair to sign anything in the wild, Nostr already has the tooling and apps, and the signed Nostr json events can reference the atproto DID as a tag.

Just link a Nostr keypair and done. It'd be so easy.