I try to imagine how Bluesky looks after it scales up to Twitter size. I discuss how aspects of its architecture lead to big differences in how far different elements can diversify, and how this leads to predictable outcomes of how they respond to commercial pressure. I judge the overall result as feasible, and better than the status quo, but complicated and fragile. Finally I suggest how cooperation with the Fediverse could look, and how that is a more promising outcome.

Bluesky has solved its scaling problems, making a DIY server viable for hobbyists. Or… has it? No, no it hasn’t, and I see quite some dangers ahead. Overall, I conclude that Bluesky’s ambitions are surprisingly narrow for a company with that much money, and it poses much less of a competitive threat to the Fediverse than you’d think.

When you block a user, it’s possible they may be notified. This is bad. Here I explain how this is possible, and the good reason why it was deliberately built this way.

How can we create a funding model that allows startup founders to build terrific products that people love on the open social web? My answer: let’s not do that.

How to migrate a Wordpress instance with ActivityPub followers using the export and import tools. And, why that doesn’t work.

Arguing that at least some ActivityPub servers should act more like email servers, allowing clients to communicate data that doesn’t fit into the server’s data model.

How you can promote events with the new Event Bridge Wordpress plugin. I think this is a model for helping the Fediverse live up to its potential.

I wonder about the definition of “decentralisation”, and realise that the political meaning of the term is far more important than the technical or mathematical senses preferred by Fedinerds.

People panic that the latest Twitter wave is going to BlueSky not Mastodon. But they are chasing different “wins”.

The Matrix Board was elected, overwhelmingly white male. It’s widely understood that this needs to be fixed. I propose some actions.