Results 149 comments of Eugene

Looking at the linked issue above - not sure if you want to defer these patches to another PR, but should also check (ideally at RPC level) that: - the...

Just a note that for zero-conf channels they will be auto-rejected unless an acceptor is present

Doesn't have to be in this PR but we'll need to update our fuzzing harness for the new message type

Is there a follow-up PR incoming that I could look at to see how this would fit together? Otherwise, imo it's hard to compare the original code vs this.

This is also needed for when we upgrade to v3 transactions or if imbued v3 becomes a thing

With imbued v3 (new v2 transactions having v3 logic applied to them), we won't need to update btcwallet or btcd for the v3 logic to apply; it will just happen.

> Much better control IMO, if we have a degree of confidence that at x rate y% of peers will accept the transaction and then y can be a configurable...

On second thought I think keeping the moving window is good because if we suddenly lose the majority of our outbound peers and only have one or two peers 1000...