Matt Corallo

Results 964 comments of Matt Corallo

> The change is already in Persist interface docs? Uhhhhhh, right. Not sure what I meant. > This is because current implementation of MonitorUpdatingPersister depends on this "updates are sequential"...

Surprisingly, with the utils, no one has actually complained about this, so removing the milestone entirely.

For electrum/esplora clients, to implement this, we probably need to think about switching the order from best_block_updated -> transaction confirmation to transaction confirmation -> best_block_updated.

That's pretty unrelated - that's just about when our counterparty is stale/has reverted to a previous state. We actually already do that. This is more about us having the latest...

Sadly I don't think we have the bandwidth in 0.0.119 to dig into this, but hopefully its mostly a documentation thing in a release soon.

We may actually just change the spec here - its not clear that this accomplishes anything in the spec aside from complication - https://github.com/lightning/bolts/pull/972

I'd let this one sit for a bit. There continues to be some debate in the BOLTs as to the correct direction here, with multiple alternative PRs going different directions.