On 31/12/2016 11:20, Masataka Ohta wrote: > Michel Py wrote: > >>> Brian E Carpenter wrote : >>> Which is exactly why we have to make multi-prefix multihoming work out of the box. Which is why we >>> have RFC 7157 and RFC 8028 and draft-ietf-rtgwg-dst-src-routing, and we're not done yet. >> >> We have missed the delivery date by 14 years. > > I wrote: > > https://tools.ietf.org/html/draft-ohta-e2e-multihoming-00 > > in April 2000 and I know it is stupid to use source routing for > multihoming. So, never deliver it. > > All we need is transport/application layer capability to treat > multiple source and destination addresses and implementations > of such TCP has been available since 2003 or so, more than 14 > years ago. I am very suspicious of any claim about "all we need" in this area, but I have written about this in some detail elsewhere: http://www.sigcomm.org/node/3484 Brian