Closed ziggie1984 closed 2 months ago
PR was tested by the person who run into this problem, and it worked for him.
Hey Ziggie. Thanks for the fix. But is it possible you're using an old base? I fixed this in 83aa752d, which just hasn't been in a release yet. Can create a release if that helps.
True, always a step ahead :).
I think it makes sense to release a new version with this bugfix.
Fixes a bug where we would assume outputs are unspent although they were not.
Example:
while sweeping the remote close of this channel:
https://mempool.space/address/bc1q96vq7jhl552ucgy79q7zqny27jmlgq6nmspmcvuw97dgsdedlpusnxamhj
the
to_remote
addr has unfortunately two outputs of two different channel closures (prob. some kind of LND bug), so when querying for the unspent addr utxos, we need to make sure before including the utxo we query the mempool for the spendness of this UTXO.to_remote addr has two channel closures associated with it:
https://mempool.space/address/bc1qkvtux0hk28qtn7njdjlym2pmhtq799r36yt8a4cuqs0ejm6z520sz4xkqe