I don't yet know enough about routing internals so bear with me. This may be the next step in #635 ...
I noticed when I try to pay some invoices, some intermediate nodes reply with this message:
"failed: WIRE_UNKNOWN_NEXT_PEER (reply from remote)"
This leads to a more general problem of buggy/broken/rogue nodes that accidentally or purposefully disrupt routes. Perhaps we should mark these as temporarily broken and route around?
I don't yet know enough about routing internals so bear with me. This may be the next step in #635 ...
I noticed when I try to pay some invoices, some intermediate nodes reply with this message:
This leads to a more general problem of buggy/broken/rogue nodes that accidentally or purposefully disrupt routes. Perhaps we should mark these as temporarily broken and route around?