[arin-ppml] Routing Research Group is about to decide its scalable routing recommendation

[Michel Py]

> Joe Maimon wrote:
> To be clear, I dont [want to] have any of my own conclusions,
> only the curiosity of the likely ill-informed as to the 
> thoughts of those involved in the problem space in this aspect.

There is nothing wrong in having your own conclusions IMHO, as long as
you understand that their influence greatly depends on how bold they are
and the "peer recognition" factor. The only reason I think that I am
somehow entitled to a voice in this is because I tried, and because I
failed.


> Even as the tech appears shiny and usable, I am unclear on
> how it will actually eliminate the PI v. DFZ debate that
> crops up here and there now and again.

Of course this is exaggerated, but I think that the PI v. DFZ is as sure
as death and taxes from where I stand from.

No matter how bright and optimistic Robin and other IRTF folks are (and
they are bright and hard-working on it), I (not the only one, doesn't
claim I invented it all) wrote years ago about the hot topics they are
currently discussing. I can remember the excitement.

That being said, in my wildest dreams, it will take at least 15 years to
replace BGP if the economic conditions were there, which they are not
anymore. As stated earlier, many people considered earlier on that
deploying IPv6 would be a lesser challenge than replacing BGP.


> It really would be nice have our cake and eat it too and I was
> hoping that a navigable path to there can be shown to exist.

So did I believe, 10 years ago. And I did spend a lot of time and some
personal money to write things, attend IETF meetings on my on dime, and
so on. I still have to see that navigable path. When I designed MHAP,
the design goal was 1 Billion multihomed sites. 

Just too complex.

Michel.