http://mailman.apnic.net/mailing-lists/sig-policy/archive/2014/09/msg00049.html
Interesting quote from the paper.
I've sometimes wondered if RIR's do too much, if there is inherent mission
creep to justify increasing revenue due to increasing member base.
While it is often useful to the community and at least mostly harmless, it may
reduce competition in areas that are not strictly needed to be monopolized.
I believe monopolies are good for many things, but the scope of each monopoly
should be well defined and no mission creep should be allowed.
I'm not sure for example, if 11MEUR is needed for number registry personnel
costs, that could give you 100 hostmasters with 5500EUR/month salary, in good
likelihood, we'd be able to run focused number registry with volunteers.
Inside monopoly, there is always honest belief that you are operating as
leanly as you can, because usually no organization realizes their
inefficiencies until they must to survive.
Quis custodiet ipsos custodes?
-b
I think your math is off? 11,000,000 / 100 == 110,000 / 12 == 9,166 month
right? Did you mean '200 hostmasters at 5500/month' ?
you'd likely also have to put into the mix the cost of infrastructure,
right? I'm not sure what current arin/ripe/apnic folk have deployed, I
imagine some servers (100k of gear? replaced every 3yrs?) and
routing/switching devices (2M replaced every 3 yrs), and link costs.
-chris
I'm not sure for example, if 11GER is needed for number registry personnel
costs, that could give you 100 hostmasters with 5500EUR/month salary, in good
likelihood, we'd be able to run focused number registry with volunteers.I think your math is off? 11,000,000 / 100 == 110,000 / 12 == 9,166 month
right? Did you mean '200 hostmasters at 5500/month' ?
you'd likely also have to put into the mix the cost of infrastructure,
right? I'm not sure what current arin/ripe/apnic folk have deployed, I
imagine some servers (100k of gear? replaced every 3yrs?) and
routing/switching devices (2M replaced every 3 yrs), and link costs.
we could nit-pick saku's arithmetic to death, but what would we learn?
it costs money and clue to run a good registry, news at eleven.
in '92 or whenever, when the nic contract went out to bid, rick said
he'd do it for free with some simple scripts. it's a long way from that
to where we are today, and i doubt either extreme is where we should be.
i suspect that if we threw out all the micro-management policies,
restrictions on transfers, barriers to entry for legacy and newcomers,
etc., we might be able to move significantly closer to rick's idealistic
position.
buy it would require a change of paradigm, and that usually requires a
lot of folk retiring. so to repeat/paraphrase what i just said in the
apnic forum,
someone too shy to post here (yes, virginia, there are such people:)
suggested i shill for them. i think their points are worth it.
reasonable public resource governance practice would include at least
the following:
- term limits for board and committee positions (maybe 2-4 years?)
- ten year employment caps on executive staff
- members decide bylaws and budgets
and as i suggested to arin, a gov/ops review consultation consisting of
folk with some stature in these areas, and not having any members from
board or staff.
i would love to see some folk with enable on the board, such as you.
but without the paradigm shift, it would just be pain and torture to no
real avail. and good folk with enable are too busy enabling the
internet as opposed to making careers as wannabe micro-policy wonks.
randy
I'm not sure for example, if 11GER is needed for number registry personnel
costs, that could give you 100 hostmasters with 5500EUR/month salary, in good
likelihood, we'd be able to run focused number registry with volunteers.I think your math is off? 11,000,000 / 100 == 110,000 / 12 == 9,166 month
right? Did you mean '200 hostmasters at 5500/month' ?
you'd likely also have to put into the mix the cost of infrastructure,
right? I'm not sure what current arin/ripe/apnic folk have deployed, I
imagine some servers (100k of gear? replaced every 3yrs?) and
routing/switching devices (2M replaced every 3 yrs), and link costs.we could nit-pick saku's arithmetic to death, but what would we learn?
it costs money and clue to run a good registry, news at eleven.
sure, my point wasn't really that 'math is wrong', so much as 'running
a registry likely costs some cake in gear/bw/admin-time'
and that i'm not sure that 11m is off as a number close to the scale
of the cost/problem.
i would love to see some folk with enable on the board, such as you.
can't other officer from same company already serving, phew! 
(see
bullfighter turn stance)
but without the paradigm shift, it would just be pain and torture to no
real avail. and good folk with enable are too busy enabling the
internet as opposed to making careers as wannabe micro-policy wonks.
I also agree that 'lots of policy' hasn't really gotten us anywhere 
I also agree that 'lots of policy' hasn't really gotten us anywhere
< cheap shot >
this is not exactly true. we just don't like where it has gotten us
randy
that's a fair cheap shot.
I also agree that 'lots of policy' hasn't really gotten us anywhere
< cheap shot >
this is not exactly true. we just don't like where it has gotten usthat's a fair cheap shot.
randy
in '92 or whenever, when the nic contract went out to bid, rick said
he'd do it for free with some simple scripts. it's a long way from that
to where we are today, and i doubt either extreme is where we should be.i suspect that if we threw out all the micro-management policies,
restrictions on transfers, barriers to entry for legacy and newcomers,
etc., we might be able to move significantly closer to rick's idealistic
position.
It is true that we have accumulated several decades worth of policies
and assumptions that originate from the baseline that "IPv4 is a very
limited resource"; whether these make sense to carry over to a world
where RIRs are not doing IPv4 issuance but merely keeping track of the
present address holder is an excellent question. There is nothing that
inherently prevents a change in approach other than the policy developed
by the community. Given that anyone can participate in the policy
development process (with results that are based on participant support
for various proposals), does the lack of change reflect just a general
lack of interest in making that happen or is it a reflection of the
hysteresis built into the system? (i.e. the perceived need to learn
the policy terminology and the policy development process, write up
a proposal, educating others in the problem your trying to solve,
participating in the discussion, etc.)
I have had folks tell me that it doesn't appear worth the effort to
change policy when all they really want is to get some address space -
that a fairly hard situation to address; how does one have policy which
truly serves the needs of entire community, when the actual participants
are volunteers and thus a self-selected subset by definition? In the
ARIN region, the presumed answer to this question is the ARIN Advisory
Council, which shepherds the development of policy proposals into draft
policies and proposes them for adoption before the community (which does
reduce the amount of process issues that someone with a good idea needs
to know in order to raise it for consideration...) It still ultimately
comes down to the show of support for a given policy proposal at the
ARIN Public Policy Consultations (PPCs) which are held during NANOG and
the twice annual ARIN Public Policy Meetings. Those who participate
(generally between 50 and 100 folks depending on meeting) determine via
their show of support whether draft policies ultimately get abandoned or
adopted. If there is an "unserved" segment of the community out there of
any size, it simply takes attending either on-site (or remotely) to be
included; making significant policy changes does not require membership,
agreements, seats on the AC or Board, or anything else other than actual
participation in the process.
buy it would require a change of paradigm, and that usually requires a
lot of folk retiring.
I'm going to disagree since most of concerns you cite above (e.g. transfer
restrictions) are set in policy, and as I indicated, that can be readily be
changed if even a small number of people got involved with a clear intent to
change them.
so to repeat/paraphrase what i just said in the apnic forum,
someone too shy to post here (yes, virginia, there are such people:)
suggested i shill for them. i think their points are worth it.
reasonable public resource governance practice would include at least
the following:
- term limits for board and committee positions (maybe 2-4 years?)
- ten year employment caps on executive staff
- members decide bylaws and budgetsand as i suggested to arin, a gov/ops review consultation consisting of
folk with some stature in these areas, and not having any members from
board or staff.
Despite the fact that I do not believe that current policy development is
encumbered by the practices you cite above, I do believe that any member-
based organization should periodically look at its accountability to the
community served. There is some commonality of belief in that principle
among the other RIR; in fact, the RIRs (working via the NRO) recently
completed a RIR survey and published a matrix providing an overview of
the governance frameworks of the RIRs. It is designed as a reference
for the global Internet community, and provides a structured overview
of various aspects of RIR governance, with links to the source documents
on the respective websites of the RIRs -
https://www.nro.net/about-the-nro/rir-governance-matrix
I do not believe that the current RIR governance matrix gets to the level
of detail necessary to show each RIR's compliance to your "reasonable public
resource governance practices" listed above; I will propose that it be
updated accordingly as a first step in this process. While everyone may
not agree on what constitutes best practices for RIR governance, there is
no reason not to have clear documentation of the current state of affairs
to aid in the discussion.
Thanks!
/John
John Curran
President and CEO
ARIN
No. I live in socialist country where every 1 gross EUR paid to employees
costs about 1.7 EUR to employer.