Jack Dorsey Unveils Bluesky Social
ssaurel.medium.com
ssaurel.medium.com
"Imminent"? sounds like this is an announcement of announcement. For those, we can wait until the actual announcement is available and then apply beta reduction.
https://hn.algolia.com/?dateRange=all&page=0&prefix=true&sor...
In the meantime, the post about the information that came out a few days ago was discussed here:
The AT Protocol - https://news.ycombinator.com/item?id=33252108 - Oct 2022 (99 comments)
Bluesky was an initiative launched by Jack Dorsey while he was CEO of Twitter, and it was intended to develop a set of federated social media protocols and eventually move Twitter to those protocols. (https://www.cnbc.com/2019/12/11/twitter-ceo-jack-dorsey-anno...)
He hired a development team for it (led by Jay Graber) who still work on the project today, probably as employees of Twitter? Jack's role in the project is unclear, apart from being a board member of the organization.
I have no idea what the relations between all the parties are at the moment, considering Dorsey x Twitter x Musk x everything else that has happened since December 2019, but as far as I can tell Bluesky – while working independently – is still fully funded by Twitter. It is not Dorsey's company as everyone seems to believe.
I've spent less than an hour reviewing atproto. So far it has passed my sniff tests; seems legit decentralizing. But there's definitely room for gotchas I havent seen yet.
Thanks for detailing where the powers lie. So far, I'm still willing to attribute great intent & purpose. But it's important to consider, to know influences. Your post is much appreciated, evem if, for now, I believe it all, ascribe- hestinatly- to this system.
https://twitter.com/bluesky/status/1582437529969917953
@dang, could we get the link changed to point to the original thread?
But I'll settle for anything that isn't a medium.com link, particularly a "member-only" post. Fuck that noise.
The AT Protocol - https://news.ycombinator.com/item?id=33252108 - Oct 2022 (99 comments)
They have recently distinguished the project proper from the "Bluesky community," which was -- as far as I could tell -- just a chatroom with a Matrix/Discord bridge, vaguely sanctioned by but distinct from Twitter. I couldn't understand most of what was under discussion, but it's a moderately amusing space if people arguing about what buzzwords mean is amusing to you.
It doesn't mention Blockchain, nor NFTs, nor web3, so plus points for that.
Looks interesting in general. The architecture looks quite nice, with a lot of separation of concerns (like we have in the www).
TLDR is that name resolution, storage, indexing, and recommendation algorithms can all be provided by separate entities, with a sprinkle of cryptography to make it possible. Will be following this to see if anything comes out of it.
I’d like to dig in more this weekend, it does seem like a pretty nice synthesis of other ideas and if it has major backing + not being attached to crypto I think it could have more longevity and adoption than other distributed web attempts.
https://atproto.com/guides/identity#did-methods
That said, and while I have serious doubts about it, they are seemingly approaching with an open mind. Unlike others they aren’t just shoving“blockchain” into every part of it so they can grab VC money and run.
If you go to bsky.app it has been “unveiled” that there’s a form to give strangers your email address to notify you when it’s been uh, double-unveiled or whatever.
I'm not saying that any of the centralized networks are immune from those maladies, but the problems seem to be substantially less challenging on that front.
So is Twitter and Instagram and more. Bots and ads. Just peruse fitness channels anywhere for the most blatant shilling you've ever seen.
That's actually the mainstream opinion outside politically-charged/weird nerd circles.
So it's like mastadon but with better security/privacy?
This isn't the sort of balance we need.
99% of people don’t care enough to put up with the complexities that decentralization brings.
So this ain’t the killer.
And those who are called killer vanished without a hit.
Remember Vero, the FB and Insta killer?
The protocol they developed (https://atproto.com/) seems to differ with ActivityPub at least in "Algorithmic Choice", a "market of open algorithms". The other features are already offered by ActivityPub - which is what Mastodon, Pleroma, GNUSocial use.
I wish they had built on top of ActivityPub instead of making their own federated protocol, but they may have some particular reasons as to why.
Why did you need to invent xrpc/lexicon? (which is not xml-rpc by the way) but a purposely leaky a schemafull RPC abstraction over HTTP with introspection.
Coulda honestly just used an real xmlrpc server at that point. They borrowed every concept except made it JSON. If nothing else it would have given you pre-built libraries for every language ecosystem.
I personally don't think this will rise that quickly(for many reasons, the main 2 being the focus on web3 and the second being jack's past), but the fact remains that the market for a non-censorable or at least trustworthy social media platform exists.
i’m not sure that it _actually_ does beyond extremely niche. any platform where it is impossible to censor would quickly be overrun with dick pill and millions of other forms of spam.
Give me control over the algo that filters my wheat, and let the chaff flow in abundance.
https://moxie.org/2022/01/07/web3-first-impressions.html
What it’s really about though:
Maybe this is the Year of The Web3 project.
Meh
Maybe it's intended to become more like the vision behind Google Wave of old?
Lol, what a joke
Looks like a generic smartphone to me.
Is this some kind of Apple consumer faux pas?
Also not sure why they had to invent a new protocol from scratch?
Am I missing something?
The people that love it, love it not for its ease of use. But because it's OK for being decentralized and you can (more or less) have one pane of glass for your tech news, social media, and barking mad conspiracy forums filled with crazy people, degenerates and war criminals.
If anyone thinks this sounds like an endorsement, don't talk to me ever.
A blockchain is a 'not crazy' potential solution?
One that sticks out in particular is a time service. I don't know who April is, and I don't know if you're April. All I know is that someone claiming to be April sent a message intended for an alleged Barbara at 11:34:23 PDT this morning. That's all anybody trusts me to do, and that's all I claim to do.
We could divorce identity from these chat services in a similar way, but we keep making the mistake of trying to positively ID people like you would for a court case, and that's not what's needed most of the time. All we need for the happy path is a way to verify that you are the same person who claimed to be this identity a month ago, instead of someone impersonating you.
The problem is that if it's free to create identities, then people can use it to do astroturfing or false flag situations. Vetting is a big place where things fall down because vetting and adoption fight each other in an arena called Instant Gratification. Also we haven't instilled the same sort of Stranger Danger on the internet that GenX and GenY got spoonfed about public spaces since they were old enough to talk.
Search I get, decentralized search is hard without centralizing because you can’t even know all the Mastadon instances that exist. Just like email Mastadon is open federation.
I suppose what Mastodon unfortunately misses is that "microblogging" ain't like email: it's supposed to be done publicly -- so, unlike email where you're likely to have a small group of people who mostly know each other -- immediate identification of "who's talking" is far more important?
Isolated communities that hear about each other by diffusion is not a bad idea.
Technical details: https://blueskyweb.xyz/blog/10-18-2022-the-at-protocol
Developer documentation: https://atproto.com/guides/overview
> Bluesky is not a blockchain, and we believe the adoption of social web protocols should be independent of any blockchain. [0]
> We’re not describing what we’re building as a federated or p2p network, or as a blockchain network, because it doesn’t fall neatly in any of these categories. It could be described as a hybrid federated network with p2p characteristics, but it’s more descriptive to focus on the capabilities – self-authenticating identities and data – than on network topology. [1]
In short, it is NOT a "blockchain social network" and doesn't even use one to begin with.
[0] https://blueskyweb.xyz/blog/2-7-2022-overview
[1] https://blueskyweb.xyz/blog/3-6-2022-a-self-authenticating-s...
Not many keywords immediately turn me off from a project, but this is a surefire one
(Which is basically web3 more or less funnily rebranded)
I think they're spot on with needing control over your identity, and them being transferable. Being locked into <username>@host.com is a big pitfall of email.
I've been working on a similar federated protocol awhile which integrates Hal Finney's RPoW concept. Moderation isn't federated which is where it diverges from things like mastodon and bluesky.
Instead, you're paid in messaging tokens by the sender to keep spam off the platform. Tokens are free, assuming you don't need mass amounts of them to spam.
I'm hoping to find some other engineers that might be interested to help work on it. Can contact me Stamp at lotus_16PSJPAVocAM5behRWxqwQnpEVRPJrV4XxbthBhJR. (Human readable handles using DIDs are something we need to get working.)
I guess this means you need to store some sort of key that can be used to create a persistent id across servers. Which is great. But we need an explicit mechanism to avoid people having to trust a centralized authority to store that key. Storing on their own devices is going to inevitably result in people losing that key.
I don't know if secret sharing can ever be dumbed down and simplified that the average non-tech person can use without getting confused.
Also lack of free speech isn't the problem with Twitter, nor do federated services solve it (control over your identity is meaningless if noone will let you distribute your content). The problem with Twitter/social media is that it amplifies all of the negative aspects of social power.
Moving that to a federated protocol won't make the environment less toxic. There won't suddenly be less misinformation (on one side) or less censorship (on the other). Social media is rotten at its core and just shouldn't have a place in polite society.
The fact that Twitter is so popular/necessary among journalists is a feature -- the profession moved from working class to being a tool/playground of billionaires decades ago and the combination of publications and social media is how they enforce their worldview on masses. The most social-media savvy journalists (e.g. Carlos Maza, Taylor Lorenz, etc) come from ultra-wealthy families.
Jack Dorsey either fundamentally misunderstands that or is complicit and doesn't care.
If people really want "internet freedom", then we're going to have to fully embrace balkanization and be prepared to run completely parallel networks with separate root DNS and compelling content to make people use it. It's a pipe dream.
That's the reason for the RPoW tokens, otherwise the platform would collapse from spam/DDoS.
Also, where you accept your DMs can move relatively quickly with the way the system is designed.
> our friends at Bluesky announced their application protocol for building decentralised social media called AT. While not based on Matrix, there are some parallels, and some stuff we may be able to get inspiration from around portable identity :) https://blueskyweb.xyz/blog/10-18-2022-the-at-protocol
Sounds like some there are some parallels, and the protocol might inspire portable account identities in Matrix!
The only account migration tool I know of at the moment is proprietary [1], I'd love to see portable identity in Matrix come to life.
Well, fortunately Matrix is working on adding OpenID Connect support[0], so I guess the answer will soon be "Matrix identities are as portable as OIDC identities".
Unfortunately, however, that means "not very portable" in practice, but the OIDC people at least recognise that fact and are apparently working to fix it with an extension called SIOP (Self-Issued OpenID Provider).[1]
[1] https://openid.net/specs/openid-connect-self-issued-v2-1_0.h...
This does not make any sense. The problem with social media isn't just "algorithms", it's their entire designs. TikTok's swipe-based UI and powerful editor incentivizes short-form video content that builds off each other. That creates certain social behaviors unique to TikTok. Another example: Twitter's decision not to have a "dislike" buton has a huge effect on how tweets are boosted. You can't "manage" that no matter what decentralized protocol you come up with, because that's baked into the UI.
Not to mention, what does it even mean to "choose your algorithm"? Once you give me your data, you have no control over what I do with it. The only thing you can really choose is what piece of data you give away, but the article as it stands doesn't really make sense.
This would mean your comment about the UI is more up to the user/client app, so this stuff is up for conversation. Unlike tiktok.
How does it use blockchains and cryptocurrency? Is this an integral part of the distributed protocol, or is it more like Mastodon?
The app that currently owns the attention of the world's youth is controlled by the CCP, and that's not going to change if we keep building apps no one wants.
Bad opening line. Hard pass.
There is a big opportunity coming for whoever is prepared to catch the foreseeable wave of twitter account mass-exodus which will take place as soon as Musk takes control and begins asserting themself.
Which tw inspired platforms are baked and ready deal with the influx?
Bold claim. Given how fast it's all falling apart I fail to see why and how.
I am excited about what Nostr[0] is doing is instead. Looks much simpler way to approach the problem.
However, it's been almost a year since Dorsey was CEO even though he didn't step down from the board until May. That is pretty close to the limit of what California would really allow in a non-compete.
Source: California law: https://leginfo.legislature.ca.gov/faces/codes_displayText.x...
> [...] any owner of a business entity selling or otherwise disposing of all of his or her ownership interest in the business entity [...] may agree with the buyer to refrain from carrying on a similar business within a specified geographic area in which the business so sold, or that of the business entity, division, or subsidiary has been carried on, so long as the buyer, or any person deriving title to the goodwill or ownership interest from the buyer, carries on a like business therein
Practical example: one customer was a cruise line, and the company had to tactfully ask if building a similar product for a hotel didn't violate the contract. As opposed to, say, a school which would not reasonably compete with cruise lines - Semester at Sea notwithstanding. If you don't want to agree to this, you don't need to sign a contract not to work with competitors. But of course, the customer is probably not going to want to pay as much if they know that they can very quickly lose whatever edge they're getting.
[applause.flac loop=1000]
It kind of feels like Dorsey wanted a blockchain, but their engineers found that blockchain-based social media did not work well and just built a federated network, so now they're going for some "inspired by the blockchain" angle.
Edit: I looked at the github, not the medium page, I see I'm wrong. It is some web3 BS
Where exactly does it say that BlueSky is yet another web3 / blockchain project?
The Medium post isn't from the BlueSky team, its from a random crypto blogger and even he doesn't reference where BlueSky is a blockchain or crypto token platform.
> We’re not describing what we’re building as a federated or p2p network, or as a blockchain network, because it doesn’t fall neatly in any of these categories. It could be described as a hybrid federated network with p2p characteristics, but it’s more descriptive to focus on the capabilities – self-authenticating identities and data – than on network topology.
From: https://blueskyweb.xyz/blog/3-6-2022-a-self-authenticating-s...
Not that it clarifies anything for me. It feels (I'm speculating) like they are very careful not to mention web3, but maybe that's because it really has nothing to do with tokens and whatnot and is actually just a protocol that lets people run their own social network servers. But it doesn't say it's that either. Some plain language would help make it clear if it's legit or just some new take on blockchain scamming
It seems like it models your data as a Merkel tree log, signed by an identity key that can be obtained through DNS. Unless they plan to add a namecoinish thing for identities you shouldn't really need Po* to "secure" that.