> In any case very welcomed changes! :-) Thanks! > From reading the diff, it looks like you know what you are doing, but the > Tracker code base seems way too complex for what it's doing. Yeah, I'm afraid it's getting complex. Basically, there is a complex trick involved only to get monitoring kicking in when a node will exist in the future, and not forget about that eventual, potential (might never exist) future node. If we could start monitoring on imaginative nodes, it would be much simpler, but it would only move complexity elsewhere I'm afraid. Philippe PS: will fix those style issues ASAP, after some other bug I'm looking into. Thanks for reviewing!