need to handle getticketsvotebits errors better

[jolan]

votebits errors out when ticket doesn't exist due to mempool differences or when a missed/revoked ticket is present or not.

[mverrilli]

This isn't only for missed and revoked tickets right? Or is this a separate issue?

I assume this is where the ticket list comes from? Could we union the responses from the different wallets since mempool could literally be any subset of the overall mempool? (Seems like there is still always a chance that some tickets may not be in the list when they are still in the mempool).

Or am I way off?

[jolan]

I was just listing some examples of de-synchronization from past issues and the goal would be to have the code work in a way where the inconsistency isn't a problem.

Can run ticketsforaddress and getticketsvotebits against each server individually.

Or just have getticketsvotebits ignore passing of an invalid ticket and work with a single response, they will eventually be consistent but then users may complain about missing tickets.

[jolan]

Bumping milestone to 0.7.0. PR #32 handles this but it needs more testing and it depends on the new setticketsvotebits functionality in 0.6.0. So once 0.6.0 is released, we can test/merge #32 and require 0.6.0.

[chappjc]

I won't be offended if PR #32 goes to the scrap heap. There's no extended bits handling, and it's complex.

[jolan]

The situation was vastly improved by #60 so I don't see a reason to keep this open.