Instead the handler should retry a connection.
Depends on D10834
Details
Details
Set a flag, dispatch connection issue and check if a user won't get logged out.
Diff Detail
Diff Detail
- Repository
- rCOMM Comm
- Lint
Lint Not Applicable - Unit
Tests Not Applicable
Event Timeline
Comment Actions
Is this what we want to do for all connection issues, even for my keyserver? Or are there some kinds of connection issues, where if they occur for my keyserver, we'd want to log the user out?
Comment Actions
We probably should prefer treating all the keyservers the same. As for the connection issues, I'm wondering if e.g. not_logged_in_error is something from which we can recover - it is set after our socket receives not_logged_in message.
Comment Actions
I think yes, we can recover from that. To be honest, it's a weird one... it can only happen if a client attempts to connect with a keyserver socket using an anonymous cookie, but based on the client code that should not be possible. But if it does somehow happen, the solution is just for the client to attempt the login.