"ATProto does not scale wide: it's a liability to add more fully participating nodes onto the network. Meaningfully self-hosting ATProto is a risk to the ATProto network, there is active reason to disincentivize it for those already participating."

#ChristineLemmerWebber, 2024

dustycloud.org/blog/re-re-blue…

Oof. That is *not* good for the claim that BlueSky's protocol can be as friendly to decentralisation and independent hosting as the fediverse.

IMHO that's game, set and match.

#BlueSky #ATProto

"In many ways, Bluesky is speedrunning the history of Twitter."

@cwebber, 2024

dustycloud.org/blog/re-re-blue…

This is exactly the comparison I've been making. Seeing it made within the context of a careful steelmanning of BlueSky's ATProto is very much a cat-that-got-cream experience for me.

Having criticised BlueSky a *lot* over the past few months, I do want to flag a couple of things that make it qualitatively different from Titter. At least for now.

1) their entire stack is Free Code, and their protocol is documented.

Christine's analysis suggests that doesn't help as much for potential self-hosting as it might. But it remains true that a well-funded organisation could stand up a complete replacement of BS, using their code.

Not true of Titter.

(1/?)

#BlueSky

@cwebber

2) BlueSky's staff, and the people using it, have the entire history of Titter to learn from.

Which means that although enshittification is a risk to any centralised, VC-funded platform, those involved are as aware of that as the critics. An Exit to Community could be baked into their business strategy. Where ownership passes to the workers when the VCs exit, or to the community of people using the platform.

Imagine a social.coop the size of BS! Curious what @ntnsndr thinks of this.

(2/?)

3) There is strong support for interoperation, from both BlueSky engineering staff and management.

Titter was very encouraging of devs building third-party apps on top of their platform. But full interop with competitors was always adversarial.

Yet we see BlueSky Issue discussions where BridgyFed creator @snarfed works through interop bugs with BS engineers. The relationship is so cordial that @EvanProdromou@socialwebfoundation has even said he considers BS a large fediverse node, like Mastodon.social.

(3/?)

So there's a few points to consider. I'm still not setting up a BlueSky account. But I have enabled BridgeFed interop with BS, just as I'm happy to interact with Nostr accounts here through any bridges that are available.

As the old saying goes, all roads lead to Rome. For now at least, the roads of all alternative social networks at least pass through the outskirts of the fediverse.

(4/?)

Back in 2017, I said;

The Fediverse / Federation aims to eventually unite all the things that still exist into one glorious meta-thing.

socialhub.activitypub.rocks/t/…

To me, that remains the goal. Let's unite all the things!

(5/5)

Fediverse history piece from 2017: A Brief History of the GNU Social Fediverse and ‘The Federation’

In late 2016, the fediverse was reeling under the onslaught of the first two major waves of people leaving Titter for our shores.

^SocialHub

Great thread.

Another value prop of all-open-source is basic algorithmic accountability.

And I think Christine didn't give enough credit to the composable moderation stack, which I think is an important component.

(1/?)

@ntnsndr
> Great thread

Thanks : ) I tagged you in the hopes for some comment on the prospects of an Exit to Community for BlueSky, from one of the founders of the Buy Twitter campaign and Zebras Unite : ) Do you think such a campaign would have a better chance of success this time? Why/ why not?

(2/2)

@ntnsndr
> I think Christine didn't give enough credit to the composable moderation stack, which I think is an important component

I doubt @cwebber would disagree. But speaking for myself (far be it from me to put words in the guru's mouth), I think this is orthogonal to the discussion of the base protocols. Composable moderation can be and is being built on top of AP by people like IFTAS.

Frankly I'm in 2 minds about this. I can see as many risks as benefits. But it's happening.

"How did humans get along for millennia without wifi-connected kitchen appliances?"

That's the question of the century.

The Internet of Things isn't benign.

1. Manufacturers aren't satisfied with making & selling just cars & appliances anymore.

It's an additional revenue stream from selling your privacy to 3rd parties, foreign & domestic.

2. It uses up internet & cellular bandwidth & electricity at a prodigious rate. Brownouts & blackouts get more frequent.

It's driving up utility rates everywhere.

That artificially inflated demand is misused as justification to thwart a fossil fuel...

1/3

owning and driving a car was fun many years ago. but these days cars have full data collection something, therefore cars feels ichy and scary, basically not fun.

something needs to be done for this surveillance capitalism in cars.

excellent comic

i've definitely reached a point where i will do anything i can to avoid 'smart' appliances.
prolly gonna move toward actively destroying them within a few years 🤣

I have a Roomba app on my phone that tells me when my "dumb Roomba" got stuck.

Ideally I'd put that Roomba on its own vlan to separate it from the other networked devices, but I don't have the hardware or the skills to do that.

But what tradeoff am I making with the app? What data can it collect? How would I even know? Is it in the terms and conditions or is it secret data harvesting?

@silverwizard The database is in a different container on the same docker installation. I copied the script from Marius' page and only changed the instance URL and admin mail address.

On my first try, I issued new passwords for redis, db, etc. Same problem. I then used the script verbatim, to make sure that I didn't introduce and password issues.

All containers created by the script have separate IPs.