[ppml] Proposed Policy: 4-Byte AS Number Policy Proposal

Geoff Huston gih at apnic.net
Wed Dec 14 00:14:39 EST 2005

my advice is to stick with something simple that does not assume too much. 
Its just attempting to describe a nomenclature that is easy to use, and 
allows AS numbers to be readily recognized (and remembered).


At 03:57 PM 14/12/2005, Marshall Eubanks wrote:
>I would not prefer hex. How many different ways do we need to
>describe 32 bit numbers ? Are we going to
>mix  hex and decimal with old and new ASN numbers ?
>I do like the <second 16 bits>.<last 16 bits> notation if there is
>going to be any aggregation of ASN numbers, which it seems like there
>will be. In that case, though, why not use CIDR notation ? It seems
>like the  proposal
>includes giving each RIR a /16 ASN block - I wonder if geographic
>based addressing models will start asking for ASN blocks (say, a /15
>to give one ASN for each US zip code) ?
>Marshall Eubanks
>On Dec 13, 2005, at 2:48 PM, David Kessens wrote:
>>On Tue, Dec 13, 2005 at 08:14:58PM +1100, Geoff Huston wrote:
>>>my mistake - I did mean 16 bit separation - I just didn't type it!
>>>ok - unless I hear to the contrary I'll run with the notation
>>>e.g. 10.1
>>I would prefer hex notation for the numbers. When numbers get larger,
>>the ASCII version gets longer and longer. Hex could compress it's size
>>considerably and we might even want to skip the '.' altogether.
>>David Kessens
>>PPML mailing list
>>PPML at arin.net

More information about the ARIN-PPML mailing list