+ /// In LN, output claimed are time-sensitive, which means we have to spend them before reaching some timelock expiration. At in-channel
+ /// output detection, we generate a first version of a claim tx and associate to it a height timer. A height timer is an absolute block
+ /// height than once reached we should generate a new bumped "version" of the claim tx to be sure than we safely claim outputs before
+ /// than our counterparty can do it too. If timelock expires soon, height timer is going to be scale down in consequence to increase
+ /// frequency of the bump and so increase our bets of success.