Re: [tor-dev] Revisiting prop224 overlap descriptor logic and descriptor lifetimes