If you were chased here from Twitter, it's just possible that you won't be aware that a lot of people are now leaving Twitter (or X if you must) for BlueSky. I mention this because I personally very much valued having a thriving maths community on Twitter while still being part of the wider world, so to speak. My experience here, which may be as much my fault as Mastodon's, is that there are lots of interesting people but not quite the atmosphere that I liked on Twitter. It would be a bit cheeky of me to suggest deserting Mastodon, but perhaps you might like to consider trying out BlueSky -- I would be very happy if we could use it to recover what Musk stole from us.
@wtgowers - thanks for pointing this out. I miss you!
What do you know about the corporate structure of BlueSky and the guardrails against it eventually getting "enshittified" as pressure for stock returns forces the extraction of value from its users? I put a huge amount of work into posts on math and physics. Having been burnt twice, by Google+ and Twitter, I will not again spend my time contributing value to a corporation that can eventually go bad. I've decided the "good atmosphere I liked at Twitter" was like the cow enjoying hay before the slaughter.
Btw, speaking very egotistically now, if you want to read my stuff in occasional concentrated doses, you can read it here:
http://math.ucr.edu/home/baez/diary/
I put my stuff there so I have it under my control no matter what happens.
Cory Doctorow has written about Bluesky and enshittification
https://pluralistic.net/2024/11/02/ulysses-pact/#tie-yourself-to-a-federated-mast
@jsdodge @wtgowers - Thanks! I probably read this once before but had forgotten most of it. I think @pluralistic says it just right:
"I appreciate that the CEO of Bluesky, Jay Graber, has evinced her sincere intention never to enshittify Bluesky and I believe she is totally sincere:
https://www.wired.com/story/bluesky-ceo-jay-graber-wont-enshittify-ads/
But here's the thing: all those other platforms, the ones where I unwisely allowed myself to get locked in, where today I find myself trapped by the professional, personal and political costs of leaving them, they were all started by people who swore they'd never sell out. I know those people, the old blogger mafia who started the CMSes, social media services, and publishing platforms where I find myself trapped. I considered them friends (I still consider most of them friends), and I knew them well enough to believe that they really cared about their users.
They did care about their users. They just cared about other stuff, too, and, when push came to shove, they chose the worsening of their services as the lesser of two evils."
@johncarlosbaez @jsdodge @wtgowers @pluralistic Lots of talk on Mastodon about being "locked in" to Bluesky when Bluesky is an open source ecosystem that the developers have made clear can be run independently of the public benefit llc. The team had been very explicit that they are trying to ensure the software isn't locked into the whole of the pb llc: https://bsky.app/profile/pfrazee.com/post/3laujhn5lfs2p
@redbassett @johncarlosbaez @jsdodge @wtgowers @pluralistic This post goes into some detail to argue that while they do seem to be aiming at decentralization, as of today key parts of the Bluesky infrastructure are not effectively decentralized.
@redbassett @johncarlosbaez @jsdodge @wtgowers @pluralistic I still hold out hope that it may become truly decentralized in fact, but it seems like there's lot of confusion between the intent of the design and the current state of affairs.
That being said, what they can fairly claim now is that there are 3rd party (mostly self-hosted, I think) Personal Data Servers (PDSs) that hold all the information relevant to an account, so if you use one of those you do really have independent control of your data. By contrast, on Mastodon you need the server instance you use to remain up and cooperative to move your account to another or export your data.
I've been thinking that what would be useful would be for someone to do a sort of Failure Modes and Effects Analysis like you might do in certain fields of engineering, i.e. to contemplate the different sorts of failure modes of each service that might be likely (e,g, enshittification, organizational collapse, intrusion) and what the impact would be on users (including any path to recovery). I think that would be more salient than a lot of discussions I see that focus directly on technical aspects of architecture.