WKBI #586, Redploying most of 127/8 as unicast public

Thu Nov 18 16:59:42 UTC 2021

Dave Taht wrote:
> I am sad to see the most controversial of the proposals (127/16)  > first discussed here. > > Try this instead? > > 
https://datatracker.ietf.org/doc/draft-schoen-intarea-unicast-lowest-address/ 
 > > >
in my mind, has the most promise for making the internet better in the
> nearer term.  > > Could I get y'all to put aside the 127 proposal and read that 
over, > instead?

/30 now becomes 2 usable IP addresses for the customer.

/29 "5 static IP addresses" now becomes 6?

Doing vrrp for a customer /29 gets a bit easier.

> > > ... > > It's ok, I'll wait... > > ... > > There were two other 
proposals concerning 240/4 and 0/8 also worth > reading for their 
research detail and attention to history.

Good thing they werent worried about wasting the IETF's valuable and 
precious time running the internet, because the masses cant possibly be 
trusted.

> > The amount of work required to make 240/4 work in most places is now 
 > very close to zero, having been essentially completed a decade ago. > 
240/4 and 0/8 checking is not present in the SDN codes we tried, and > 
we ripped the 0/8 check out of linux 3? 4? years back. Saves a few > ns. 
 > > All but one iOt stack we tried worked with these, many of those > 
stacks still lack, or have poor ipv6 support. esp32 anyone? > > Just as 
ipv6 today is not globally reachable, these address spaces > may never 
be globally reachable, but defining a standard for their > potential 
sub-uses seems like a viable idea. > >

On the face of it, any of the above statements should be enough to 
silence and shameface any and all of the historical naysayers.

However it would appear they are still living in the past, as evidenced 
by continued citing of "pre-exhaustion burn rate" as if the excesses of 
the past are somehow relevant to the consumption of the future other 
than an indictment of non-prudence.

Joe