Andrew Godwin

Mastodon's oauth implementation sadly tends not to be fully compliant with the spec, and I believe this is one such case: ``` 4.1.2. Authorization Response If the resource owner grants...

It's pretty much all contained in https://github.com/jointakahe/takahe/blob/main/api/views/oauth.py - we have our own oauth backend almost entirely because of the implementation being weird.

Well the problem is that we've now reached my cutoff time for keeping working on the project if nobody turns up to help full time, so there is a chance...

Yup, this is on our roadmap for approaching 1.0, though some initial work has already started.

It's not possible to export them natively, but it's possible to either crawl them all through the API (if they aren't protected), or we could support traversing a database dump...

It would be at the same time - I'd want to implement the account moving system that's already out there, so it would just let you move to/from anything else...

It's the same system - one of the reasons I have not done another release yet is that I've implemented inbound migration, but not yet implemented outbound, and it feels...

I think the problem with changing the SearchVector language is that it's embedded into the index, is it not? We can't have 20-odd indexes on the content of posts, one...

I would also err on the side of "don't put unlisted posts in a semi-public feed"

Can you provide a URL of a specific post that breaches the limit?