Matt Corallo

Merging, the diff since @arik-so's ACK is mostly just a squash: ```diff $ git range-diff -U1 7d2d04798daa9c78f183424f73f3fea6c8564573...9cc99b27e0b6471629d35df79518b7d282041e17 1d0c6c60c6802126b3b29d6a2aa026c1aa33db02...17b77e0bcf0fd6721ac820df07a92ff697b3f50f 1: eecf12b38 = 1: 73bc0f61b Add `ChannelError::close` constructor 2: 6aa0d3004 ! 2:...

Indeed, pre-Bitcoin Core 28, fee rate disagreements mean the channel is in a state where you cannot force-close it. We either have to accept that or force-close immediately. Luckily, with...

Oh, we may want to just slip this, really we want to expose a *lot* more info, ala #2618 but I haven't had the time to work on #2618 :(....

I think push it out via an update (we probably already do as a close once we shut down, or if not should?), and then rebroadcast in the monitor.

So I had a coop close transaction that timed out of default mempools, and then eventually the normal rebroadcast kicked in and replaced it with a higher feerate unilateral close,...

I believe it was via the new rebroadcast paths in 115, but I didn't check the logs. This was only possible because the original coop close (which was not marked...

Or at least have an option to tell LDK to switch what its broadcasting.

Not critical cuase we can always get our money back via FC, but its really dumb to require that.

> Check that we re-claim payments from other MPP channels if we restart and only managed to claim on one channel, specifically: This is, somewhat surprisingly, not currently broken, but...

We're not gonna get all the testing we wanted, but we'll get #3414 for 0.1 and that handles the only known-broken case. There's some less common cases described above which...