[arin-ppml] Ted's Comment on 2009-2

[Ted Mittelstaedt]

> -----Original Message-----
> From: Scott Beuker [mailto:Scott.Beuker at sjrb.ca] 
> Sent: Wednesday, April 29, 2009 10:28 AM
> To: Ted Mittelstaedt
> Cc: ARIN PPML
> Subject: RE: [arin-ppml] Ted's Comment on 2009-2
> 
> > I never said they were.  Since we provision DSL over Qwest phone
> lines,
> > Qwest is in the position of being able to -significantly- 
> undercut us 
> > on price, which they do all the time.  I know Qwest has already done
> the
> > analysis and they don't want to have anything to do with 
> the type of 
> > customers we serve.  But they have absolutely set things up so that 
> > they hoover
> up
> > the cherry customers.
> 
> Fair enough, I don't know the backgrounds of posters on this 
> list and the situations their various businesses are in, so 
> that you are in a unique situation of using their 
> infrastructure was not apparent.
> 
> 
> > As I said in the meeting comment I made, I thought this 
> policy needed 
> > tweaking.  However I support the philosophy of it.
> 
> And as I said in the comments I made, I could support a 
> policy like this if it were modified to be fair to all users 
> of IP space, regardless of size. But it sounds like you and I 
> have vastly different ideas on what the philosophy is here. 
> 
> > This isn't about driving IPv6.  It's about blocking IPv6.  
> Right now 
> > there's ISPs ready and willing to deploy IPv6 and they 
> can't - because 
> > they are single-homed and their upstream - a larger ISP - 
> isn't ready 
> > and isn't routing IPv6.  We see these complaints all of the time on 
> > the mailing lists, so I don't buy the argument that we are 
> all in this 
> > together.  Some of us are ready, others aren't.  If it's an 
> end-node 
> > AS that isn't ready, I don't care.  If it's a transit AS that isn't 
> > ready, I do care - espically when there's end-node AS's connected to
> it
> > that are chomping at the bit to get rolling.
> 
> Then might I suggest that you are barking up the wrong tree, 
> and need to start looking at your upstream options or better 
> ways to get the attention of your upstream provider. ARIN is 
> not the place to try to force a single transit provider to 
> provide a service you want, even if you feel it's 


Um, the fact of the matter is that -WE- don't have that problem since
our upstreams supply IPv6.

Others do, however.

> 
> > Clearly, the transit AS's need to get IPv6 deployed first.  
> Since the 
> > largest ISP's with the highest IP consumption rates are all in this 
> > club, and since they can't use the small IPv4 blocks that will be 
> > available after the larger IPv4 blocks are assigned anyway, it's a 
> > no-brainer to tell the largest ISPs - who need to be routing IPv6 
> > first - that they will lose access to IPv4 block requests first, 
> > before small orgs do - as we approach IPv4 runout.
> 
> As I've said, many do, and many others are just around the 
> corner. Why aren't you using your transit budget to reward 
> the organizations that provide the services you need, if you 
> need it right now?
> 
> I do not buy your argument that you can't get IPv6 transit. 

-I- am not making that argument.  As I said, we have seen these
complaints on this list.  A search of the list archives will provide
plenty of examples.  I invite you to engage with those
people and explain your philosophy of how money motivates and
listen to their explanations of why you don't know what your
talking about.

Personally, I agree with your "money talks, BS walks" argument, I've
made it myself, before.  But,
I don't presume to tell some ISP in North Dakota how their upstream
feed is supposed to act when threatened.  They tell me that they cannot
use financial motivation to force their feed to supply IPv6 - I'm
going to take what they say at face value.

But, let's get back to some realities.  As I said, the largest ISPs
will be unable to obtain usable IP number blocks from the IPv4 schema
for some time before the actual last IPv4 block is assigned.
Comcast recognizes this, which is why they made this proposal to 
begin with.  Since that is the case in reality, what is the objection to
merely lengthening the time that large ISP's cannot get IPv4 allocations
satisfied, so that we can extend the time that the smaller ISP's can
still get IPv4 allocations?