Ohta-san:
| <draft-ohta-address-allocation-00.txt>
While I agree with you that the current usage-based allocation
system is wrong, your draft's "Assignment Plan" (not more restricted)
proposes to continue an anti-market single-seller model for IP
addresses of both IPv4 and IPv6 flavours. There is no scope
for negotiating with the monopoly seller, given this tariff.
On the other hand, I do particularly like The More Restricted Assignment Plan:
No IPv4 address space should be allocated to an ISP, unless the
ISP support fully operational fully transparent IPv6 service with
at least 64K IPv6 subnets to all the end users.
because that will force IANA out of its ostrich position with
respect to being a monopoly seller with a non-negotiable monopoly
tariff that imposes significant costs upon consumers, by immediately
forcing the monopoly to stop "selling" addresses except to people
who meet extraordinarily onerous and expensive conditions.
Unfortunately, because you do not actually propose only
the More Restricted Assignment Plan, your draft effectively continues the
objectionable practice of deliberately introducing artificial
scarcity into IPv4 addresses in order to force your politics upon
ISPs and other businesses. This is identical to a monopoly which
has goods to sell but nevertheless deliberately restricts supply
in order to support higher prices.
There are two main differences between your draft's proposal and
the current system.
One of the differences is that your political ideals
include the deployment of IPv6, which is something the current
usage-based allocation system does not. This is simply a change
of the monopoly tariff, the "price" at which the monopoly will
"sell" consumers (non-transferable) address ranges. So, while
it is an important difference, it is not particularly interesting,
since it is just a higher price in view of a smaller supply.
The much more interesting difference between your draft and the
status quo is that artificial scarcity of IPv4 addresses would evolve
as a result of over-allocating IPv4 address space to applicants,
rather than attempting to allocate the smallest workable amount of
address space, as is the practice now.
This erodes IANA et al.'s monopoly supply, because now
there is a surplus held by many other parties, who then
can act as alternative suppliers of IPv4 address space.
If steps are taken to avoid the development of a massive black
aftermarket for IPv4 addresses overallocated by IANA et al., by providing
the mechanisms of a "white market" -- notably a public registry of
IP address title, with an exclusive but transferable right to
transfer title to another party -- then I would object much less
strenuously to your draft, since it is fundamentally PIARA, but
with a rather odd auctioning system for the remaining not-yet-allocated
IPv4 address space.
Given the involvement of one of your co-authors in the original PIARA
work, I am not at all suprised that the draft can easily be read
to favour the ultimate development of a market for IPv4 addresses.
Let's just not make that market completely black, with all
post-IANA/registry-allocation transactions completely sub rosa.
Note that the development of a "white market" public registry
does not rely upon the IANA. If the IANA and its registries
were to immediately cease offering IPv4 addresse AT ALL, it is
quite clear that market forces would arrive upon a suitable
solution rather quickly. Given that the initial allocation
prices proposed in your draft are extremely onerous when combined
with the conditions in RFC 2450 ("higher cost of acquiring bundled goods,
rather than only one good individually; cross-subsidy"), I imagine
that there could easily develop a situation in which IANA et al.
simply could not find a buyer prepared to meet their price / qualified
to meet the allocation conditions.
Bear in mind that the IANA IP allocation system has two functions:
1/ prevent namespace collision
2/ provide one of many inputs which network operators
may choose to use when configuring their networks
Both of these functions can easily be done elsewhere.
There just has been no reason to do that yet.
Your draft would supply a very strong reason, therefore I support your draft.
How do we get it adopted quickly, and get the IANA, APNIC, ARIN and RIPE
to IMMEDIATELY cease offering IPv4 address space to people who do
not FULLY comply with the requirements in your More Restricted Assignment Plan,
and the various RFCs and standards-tract documents it rests upon?
Sean.