(comparing packet routing to payment routing in that last email) On Mon, Jul 3, 2017 at 1:58 PM, Jehan Tremback <jehan.tremback@gmail.com> wrote: > I think proactive distance vector is less well-suited than reactive > protocols (like AODV), because > > - Links change more quickly as their balances are used up. > - Latency is not as high of a concern (nobody is going to mind waiting a > few seconds to save a few bucks). > > Just my opinion > > On Mon, Jul 3, 2017 at 1:50 PM, Tony Arcieri <tony@chain.com> wrote: > >> On Mon, Jul 3, 2017 at 7:05 AM Xavier Vas <xavier@tr80.com> wrote: >> >>> Basically, the interledger routing problem isn't different from any >>> peer-to-peer network, even if the "peers" are pools of machines on a >>> blockchain or similar. >>> >> >> There are any number of sources of precedent to look to for ILP routing. >> ILP routing protocols need to seek tradeoffs between time and transfer >> fees, something proprietary payment networks have done for many years, but >> a problem not too far off from distance vector routing, or s/OSPF/"open >> cheapest path first"/ >> >> That said, I think despite this ample precedent that what Interledger is >> doing is somewhat unique and a worthy area of exploration. >> >> I have some pejorative things to say about the routing protocols you >> referenced, but I think they're off topic for this thread. If you really >> want to know, ping me personally. >> >>> >
Received on Monday, 3 July 2017 21:23:31 UTC