This is revealing one major limitation in the whole idea of "you can move servers any time you want": it only works if the servers are still up. If the server is down, and stays down (which at this point I don't know if it's going to be the case with venera.social) you can't move. You can create a new account elsewhere, but whatever's connected with the dead account is gone.
The #fediverse needs redundancy, methinks. Some sort of service that backs up account networks.
This case with venera.social (and libranet.de) is due to a database update what went catastrofically wrong. It's been getting worse and worse for the last 11 days. And at this point I'm getting somewhat skeptical that it's ever gonna get sorted out. In normal circumstances, you'd put the backups on, investigate the problem and start over.
I was thinking that the proposal would remove the need for a cache system because you'd be in control of you. Not the instance. So if instance gets deleted, it'd be no big deal.
As far as I understood it, that's more of a tweak in the functioning of AP federation overall, closing the gap between the fediverse proper and how Bluesky works (and maybe opening way to full integration between the two protocols).
My idea doesn't require any fundamental changes. It's just a service built on top of what's already there, and I can't really say how feasible it'd be.
@BeAware It might, I guess (where'd one's data be stored? Locally on one's devices? That also can be dangerous. Distributed throughout the whole network? Hm... I dunno...), but I also suppose that you'd have a much harder time approving and then implementing something that changes stuff at the core of the whole shebang than just building a new service on top of what's working now.
Then again, that's me talking off of my derrière. I don't know nearly enough about all this to have solid opinions.