"Hyper-crawlable data", I like that 
#AtmosphereConf
(Mastodon does not like that)
Daniel talking about what if #Bluesky goes evil or gets stupid :D #AtmosphereConf
Kuba, maybe you can explain this to us because those pictures are not very comprehensible alone.
To be more specific in this question. Let's hypothesize, as the speaker does, that #Bluesky becomes evil. There are 30 million users right now everyday using the platform. What is it specifically that they do to switch from Bluesky to this Duck?
This is a presentation offered by #Bluesky, so they must be recommending something practical to their users. What is it? Or is it just nonsense presented to promote their "billioaire-proof" marketing campaign, but practically rubbish.
@mastodonmigration @mackuba IMO The two biggest adverserial use cases that BSky needs to answer are:
1.) What's to stop a product decision along the lines of Google discontinuing XMPP support?
2.) What about Relays? The extremely brief discussion in the video focuses solely on AppViews. The expression "has access to the same data that the BlueSky AppView does" is doing a vast amount of heavy lifting. The expensive part of a not-BSky is the Relay and it's also a major architectural chokepoint.
What is objectionable is this kind of tech-bro hand waving that profers solutions that they know to be impractical, probably infeasible. At the end of the day it is simply lying to their users, something which has become so commonplace as to become almost unnoticeable in the corporate tech industry.
@kittylyst @mastodonmigration Relay is just an optimization really, so you have one merged stream to pull data from, but you could have the AppView just pull from all separate PDSes directly without a relay. And you can run the relay for quite little these days. AppView is really the hard and important part.
@mastodonmigration This is a developer conference, meant to inspire devs and other builders to build things, not really user-focused
@mastodonmigration Nowhere yet - but hopefully someone watching this will make one someday :) This is a conference for ATProto developers, so the people speaking there and watching it are building all kinds of cool things on the protocol.
@mastodonmigration But a very limited actual live example: Bluesky made it so that when person B replies to person A and then A blocks B, nobody else can see that comment by B in A's thread. It's called a "nuclear block".
I disagree with this product decision, so I made it so in my thread reader tool Skythread, you can click a button to display those "hidden replies" (which are fetched from my own database partially). This is possible because I can make such independent index of all posts.
Granted there are lots of cool technical things going on, but the issue is that the company is representing something to the world--billionaire-proof, decentralized--that simply is not true. This is false advertising, no matter what they hope to someday be implemented.
@mastodonmigration It might be at least partially theoretical for now, yes
And you seen the latest analysis by @cwebber?
Her bottom line: "ATProto still cannot scale down and scale wide, not towards meaningful decentralization."
https://social.coop/@cwebber/114194820571954617
It seems, at the very least, when experts in the field are questioning even the feasibility of the technology to ever fulfill the marketing hype, it is inappropriate to continue to trumpet the ficticious capability as though it were already a reality.
@mastodonmigration I've seen those threads from her, but I couldn't find time to go through all of them and the responses in detail yet. Although, being from the ActivityPub sphere and one of its creators, she is kind of obviously biased in the direction of preferring AP and thinking from its perspective.
Nowhere. Thank you for your honesty here. Would love for that same honesty to be reflected in #Bluesky marketing.
Granted this is a developer conference and they profess to hope someone will someday invest the money and effort to make AT Protocol actually decentralized. It's like a roadmap. Aspirational.
Perhaps they should hold off in selling the platform to users as billionaire-proof and decentralized, until it is actually decentralized and billionaire-proof.