NodeBB

@catraxx @c0debabe i doubt the intended use case was to prevent reverse engineering. maybe a happy side effect, but that just looks like tailwind (or similar) & a name mangler (generally for avoiding clashes but well, not required if you're using tailwind)

@volpeon hm. i wonder how possible this would actually be with the current AP spec. i mean mastodon already has support for handles that are on domains that aren't the same as the domain where the instance resides, but idk how how implementations would react to changing the actor URI for a given handle

worst case scenario i guess it might work to just have the PDS be the "instance" as far as AP but all it does is 302 or just proxy requests to the real instance

and i feel like having multiple domains pointing to a single instance would cause headaches

@volpeon i mean isn't that just how AT&bsky works?

i think my least favourite thing about AT is how the AppView can impose its own limitations on what data it will show, but you can just skip the AppView and ask the PDS directly which will give you all the data without any restriction, yknow the whole reason private accounts don't exist on bsky

i think a more ideal approach would be for a PDS equivalent to only respond to a nominated instance, and any request for user data would be forwarded to that instance which will be able to enforce authfetch, private accounts etc