[feature] asynchronous operation

[mirabilos]

Is your feature request related to a problem ?

I’m always frustrated when I cannot favourite/boost a post or reply to it when their instance is currently having problems (e.g. parts of Hetzner are offline at the moment).

Describe the solution you'd like.

AP is often liked to eMail, so I’d expect asynchronous communication. I post a reply, my instance takes it from the client and will then later deal with the rest (and ideally already add it to the thread locally), including retrying delivery for up to two weeks, with a notification (which includes the Markdown of the reply that was attempted to be sent as well as the URL/URI(?) of the post that was attempted to be replied to) if it’s not deliverable.

Similarily, if I address multiple accounts, one of which is not currently reachable, that ought to not negatively influence the ability to post that toot and send it to those addressees who are reachable atm.

Describe alternatives you've considered.

retry until it works, which blocks the entire client, or give up

Additional context.

No response

[mirabilos]

This is only for instances, posts and accounts that are already in the local database, of course.