Re: [address-policy-wg] 2015-05 proposal

[Jim Reid]

> On 9 May 2016, at 15:16, Arash Naderpour  wrote:
> 
> 
>> Tough. We’re out of IPv4. We’re all struggling due to a lack of IPv4 
>> resources. Everyone just has to make the best of it with whatever they have 
>> now. Anyone >planning to grow their network using IPv4 simply cannot base 
>> their plans on repeatedly going to >the NCC and asking for more. It’s that 
>> simple.
> 
> No, not everyone struggling due to lack of IPv4. There is a market and some 
> are selling their IPv4 to others. Are they struggling??

I don’t know. You’d need to ask the ones who are selling. I imagine some might 
be off-loading surplus IPv4 space because they’ve switched to IPv6 or are well 
down that path. FWIW the last time I looked, it was mostly legacy space which 
was changing hands. That’s not subject to RIR policies.

Not that this matters. Yes, of course some LIRs will have extra IPv4 addresses. 
That’s largely an accident of history. The fact they benefited from the more 
liberal RIR policies from N years ago is no reason to tear up the current 
policy. Amongst other things, the current policy aims to share the pain of IPv4 
run-out roughly equally between all LIRs. 

Supporters of this proposal appear want to avoid some of that pain for 
themselves (or delay it for a while) at the expense of others. That’s not fair, 
reasonable or balanced.

>> Let’s suppose 2015-05 is adopted. We quickly burn through the remaining IPv4 
>> pool because some LIRs continue to grow their IPv4 networks instead of 
>> coming to terms with the end of IPv4. 
> 
> That assumption is not necessary valid

Nonsense.

Current policy says every LIR gets exactly one /22. 2015-05 says an LIR can get 
more than that and keep on coming back for even more. This means the pool of 
IPv4 at the NCC will be depleted at a faster rate than it would under the 
current policy. 2015-05 even says this.

> Can you please define first what you mean from "Quickly burn"? and how this 
> policy can do that?

The usual dictionary definition of “quickly burn” should be clear enough. If 
not, replace “burn” with “squander”, “deplete”, “exhaust”, “fritter away”, “use 
up”, etc.

Current policy says every LIR gets exactly one /22. 2015-05 says an LIR can get 
more than that and keep on coming back for even more. This means the pool of 
IPv4 at the NCC will be depleted at a faster rate than it would under the 
current policy. 2015-05 even says this. How often do I need to repeat this?

Suppose a boat has a leak below the waterline. What happens if you put another 
hole in the hull? Will it sink earlier or later than if it had just one hole?

>>> We need a balance between resource conservation and fair treatment
> 
>> IMO the existing policy already achieves that. 2015-5, if adopted, does not.
> 
> Really? How 2015-05 make it unbalanced?

This has already been explained many times. The proposal actually says so 
itself. I quote: "Further allocations will speed up the depletion of the free 
pool.”

A policy proposal which will speed up the depletion of the free pool BY DESIGN 
is deeply flawed and cannot hope to be either fair or balanced.

The balance in this proposal is about as “fair and balanced” as Fox News. 
2015-05 aims to encourage IPv4 exhaustion and is the very antithesis of 
resource conservation. Now let’s assume we agree that IPv4 exhaustion is a Good 
Thing. Who benefits from that policy? And are the benefits worth it? Well, the 
LIRs who’d gain from 2015-05 are the ones who can’t/won’t do anything about 
IPv4 exhaustion until it’s all gone. And even then, they’d only be able to keep 
going with those flawed models for a little longer than they would with the 
current policy.

IMO that’s not in the best interests of the community as a whole. It doesn’t 
result in an acceptable trade-off which justifies burning through the last 
dregs of IPv4.

2015-05 would allow some LIRs to keep on using IPv4 when they should have faced 
up to the reality of IPv4 exhaustion. That may well be to their short-term 
advantage. But it doesn’t change the outcome. It just brings forward the date 
when the NCC runs out of IPv4. That's
not fair for future entrants. It’s not fair to the LIRs who have already 
incurred the costs and expense of deploying NAT or IPv6 or whatever because 
“more IPv4” was no longer an option. They could have deferred or not had those 
hassles if there was a chance of getting more than a final /22.

It’s rather ironic that supporters of 2015-05 complain that they’re at a 
disadvantage to the LIRs who enjoyed the more liberal allocations of the past 
when they want to introduce a more liberal address allocation policy now which 
will disadvantage future entrants. Ho hum.

Re: [address-policy-wg] 2015-05 proposal

[Arash Naderpour]

>Tough. We’re out of IPv4. We’re all struggling due to a lack of IPv4 
>resources. Everyone just has to make the best of it with whatever they have 
>now. Anyone >planning to grow their network using IPv4 simply cannot base 
>their plans on repeatedly going to >the NCC and asking for more. It’s that 
>simple.

No, not everyone struggling due to lack of IPv4. There is a market and some are 
selling their IPv4 to others. Are they struggling??


>Let’s suppose 2015-05 is adopted. We quickly burn through the remaining IPv4 
>pool because some LIRs continue to grow their IPv4 networks instead of coming 
>to terms with the end of IPv4. At that point, those LIRs will have a very 
>nasty shock. There would >literally be no IPv4 remaining at the NCC. So these 
>LIRs will then be forced to do something else: use NAT, deploy IPv6, use ALGs, 
>buy addresses on the secondary market, whatever. These LIRs could and should 
>be doing that "something else" now. They’re >going to *have* to do it 
>eventually and might as well start now if they’ve not already done that.

That assumption is not necessary valid,  when 2015-01 was accepted we had not 
that much change in depletion rate. Can you please define first what you mean 
from "Quickly burn"? and how this policy can do that?

>These LIRs surely know today that they cannot continue with a model that 
>assumes they can issue IPv4 addresses forever or go back to the NCC and get 
>more. So all these LIRs would have achieved with 2015-05 is buy themselves a 
>little extra time to persist >with a doomed model that they already know no 
>longer works. At that point the NCC would have no IPv4 left for any future 
>entrants who will need some IPv4 to connect to the legacy v4-only Internet.

Indeed, that’s why there are supports from a part of community.

>Burning through what’s left of IPv4 for the short-term benefit of LIRs who 
>can’t/won’t face up to the exhaustion of IPv4 seems wrong to me. Future 
>entrants would not thank us for frittering away those resources. They’ll need 
>some IPv4 too. We should think >about their needs too. Our IPv4 address 
>policies can’t ignore that in favour of some what appears to be 
>mistaken/misguided short-term benefit and self-interest.


>> We need a balance between resource conservation and fair treatment

>IMO the existing policy already achieves that. 2015-5, if adopted, does not.

Really? How 2015-05 make it unbalanced?

>> in addition to those already proposed my suggestions are:
>> 
>> -allocations of a /22 every 18 months only from IPv4 Addresses Available 
>> Outside 185/8 to small LIRs (if LIR own < /20 of total allocations)

>That’s discriminatory. It’s unfair on those LIRs who have a /20 or greater.

No comment!

>> -leave the current policy only for IPv4 Addresses Available in 185/8 to 
>> avoid the too fast depletion of resources

>No. 185/8 should not be treated as “special”. The current last /22 policy 
>applies to all IPv4 blocks at the NCC. That policy kicked in as soon as the 
>NCC made its first allocation from 185/8, the last /8 it got from IANA.

>> -the addresses acquired with this system should not be able to be 
>> transferred for a long period of time to avoid profit

It is not really required to treated it as "special", when it is required we 
can have a new policy to change the depletion rate. this is not the last policy 
we can discuss here :)


Cheers,

Arash

Re: [address-policy-wg] 2015-05 proposal

[Jim Reid]

> On 9 May 2016, at 11:37, Fabio Zannicolò - Voix s.r.l.  
> wrote:
> 
> Today small companies have competitiveness problems due to lack of IPv4 
> resources.

Tough. We’re out of IPv4. We’re all struggling due to a lack of IPv4 resources. 
Everyone just has to make the best of it with whatever they have now. Anyone 
planning to grow their network using IPv4 simply cannot base their plans on 
repeatedly going to the NCC and asking for more. It’s that simple.

Let’s suppose 2015-05 is adopted. We quickly burn through the remaining IPv4 
pool because some LIRs continue to grow their IPv4 networks instead of coming 
to terms with the end of IPv4. At that point, those LIRs will have a very nasty 
shock. There would literally be no IPv4 remaining at the NCC. So these LIRs 
will then be forced to do something else: use NAT, deploy IPv6, use ALGs, buy 
addresses on the secondary market, whatever. These LIRs could and should be 
doing that "something else" now. They’re going to *have* to do it eventually 
and might as well start now if they’ve not already done that.

These LIRs surely know today that they cannot continue with a model that 
assumes they can issue IPv4 addresses forever or go back to the NCC and get 
more. So all these LIRs would have achieved with 2015-05 is buy themselves a 
little extra time to persist with a doomed model that they already know no 
longer works. At that point the NCC would have no IPv4 left for any future 
entrants who will need some IPv4 to connect to the legacy v4-only Internet.

Burning through what’s left of IPv4 for the short-term benefit of LIRs who 
can’t/won’t face up to the exhaustion of IPv4 seems wrong to me. Future 
entrants would not thank us for frittering away those resources. They’ll need 
some IPv4 too. We should think about their needs too. Our IPv4 address policies 
can’t ignore that in favour of some what appears to be mistaken/misguided 
short-term benefit and self-interest.

The NCC will reach IPv4 exhaustion point of course. The current policy ensures 
this happens much later than it would than if 2015-05 was adopted. That is a 
Good Thing. And that is more than enough reason to reject 2015-05.

> I consider unfair the current treatment, the big companies have stocks of 
> IPv4 addresses.

The holders of those resources got them as a result of the prevailing 
allocation policies at that time. Apart from the implementation detail of the 
size of the allocation an LIR gets, this is exactly the same as how requests 
from those who apply today get IPv4 addresses according to the allocation 
policies which are in use now.

BTW I can’t travel back in time to buy the ticket which would have won 
yesterday’s lottery. That’s unfair too.

> We need a balance between resource conservation and fair treatment

IMO the existing policy already achieves that. 2015-5, if adopted, does not.

> in addition to those already proposed my suggestions are:
> 
> -allocations of a /22 every 18 months only from IPv4 Addresses Available 
> Outside 185/8 to small LIRs (if LIR own < /20 of total allocations)

That’s discriminatory. It’s unfair on those LIRs who have a /20 or greater.

> -leave the current policy only for IPv4 Addresses Available in 185/8 to avoid 
> the too fast depletion of resources

No. 185/8 should not be treated as “special”. The current last /22 policy 
applies to all IPv4 blocks at the NCC. That policy kicked in as soon as the NCC 
made its first allocation from 185/8, the last /8 it got from IANA.

> -the addresses acquired with this system should not be able to be transferred 
> for a long period of time to avoid profit

Define “long period of time” and “profit”.