Yes very interesting indeed. A whole lot of work we have done in routing has been very important to ensure we retain the kademlia invariant across nodes. These are some of the hidden work recently, but of huge importance. Reading this paper enforces this hidden work is very important. So partitioning etc. is already something guarded against and will continue to be (and continue to be hidden). It’s hard though when folk scream when when launch wehn when, these papers though help when we can say - yes we seen that and have our fix or perhaps there is an issue we also must solve. This paper shows we are definitely on track though and I am glad we have taken some of the recent steps we have, although we switched off some security right now to confirm many other parts. All good though. Delay type attacks are important though and do need mitigated, for us it’s good as we are focussing on weaker nodes just now, finding them and working around them so again this is spot on.
There is a lot of good info here though for any decentralised project and I am delighted to see encrypted communications involved as well. I am surprised they did not consider even telehash or similar as a mechanism though as it seems such an overlay would help a lot.
I think agian bitcoin helps us here with such reports and hopefully we can help them with much of our work at the network levels.
Another interesting point there is restriction to /24 addresses, a neat idea but would need some real world stats to counter the banned odes verses their use etc.
All in all nice to see such papers.