[ppml] Collapsing Residential and Business Privacy (ease of use) Was: Re: Privacy of Non-Residential Reassignments in Public Whois

[william(at)elan.net]

On Wed, 19 Apr 2006, Martin Hannigan wrote:

> At 12:25 PM 4/10/2006, Divins, David wrote:
>> Due to popular demand....Attempt number 3 at an accurate Subject :-)
>
> During the XVII meeting, I talked to the author of the residential
> privacy policy, David Divens, and Aaron Hughes, regarding their
> concerns over residential and business privacy.
>
> My suggestion to the AC (and proposers) regarding
> proposals would be a rewrite to accomplish the following:
>
> - eliminate differentiation between residential and business
> - designate /29's and smaller as private

Actually I'm in favor of making /26 as minimum block required for
public whois registration. The data below confirms that this is
good boundary. I suggested similar before too...

> - reduction of NA postal codes to 3 characters

That is also fine. 3 characters is enough for US postal codes to identify
geographic area.

> - creating a confidential/undercover registration clause to allow
>   LEA to mask registrations for investigative, intelligence,
>   or other purposes as long as they identify these to ARIN
>   staff AND ARIN is able to handle such information per FISA, Title III.
>   CALEA, and other applicable regulations (IANAL). This
>   follows a concept invoked by DMV's related to license plates.
>   (and a memory jogging by Heather Skanks - thank you!)

There is too much complexity that you will impose on ARIN if you do it
this way.

I don't think ARIN is using sub-/24 allocations much for registration 
purposes so I think just increasing the boundary will be good enough.

> My recommendation is based on the following prefix distribution
> data that we have compiled based on whois data not older than
> 2 weeks. It shows that /29 is over 60% of all data and we would
> improve overall privacy by X factors. I think it is fair to say that
> the vast majority of residences are within /29, and I agree with Owen Delong
> that privacy is not an expectation for business whois data.
>
> This is more balanced than a complete masking of location data.
> I would like to hear what LEA's think of this, and I would be happy to
> consider adjustments on the confidential registration idea.
>
> Current applications of whois data include geo-location, which does
> not necessarily rely solely on whois data, but does use it for triangulation
> purposes. I think we would be surprised at the list of applications utilizing
> the postal code for this, and I am informing other geo-locators of this
> proposal and location of discussion so that they may participate if desired.
>
> MASK   PFX
>
> 4       2
> 6       1
> 7       2
> 8       188
> 9       1
> 10      6
> 11      13
> 12      36
> 13      81
> 14      216
> 15      411
> 16      7287
> 17      681
> 18      1399
> 19      3170
> 20      6004
> 21      4794
> 22      10262
> 23      19743
> 24      120053
> 25      33036
> 26      38778
> 27      103976
> 28      137726
> 29      847640  (66% of all registrations)
> 30      184
> 31      3
>
> Non-CIDR=11078
>
> --
> Martin Hannigan                                (c) 617-388-2663
> Renesys Corporation                            (w) 617-395-8574
> Member of Technical Staff                      Network Operations
>                                                hannigan at renesys.com
>
> _______________________________________________
> PPML mailing list
> PPML at arin.net
> http://lists.arin.net/mailman/listinfo/ppml