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

Geoff Huston gih at apnic.net
Tue Dec 13 04:14:58 EST 2005


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 <high>.<low>

e.g. 10.1


  Geoff

At 06:14 PM 13/12/2005, Owen DeLong wrote:
>I think it's fine, although, I'm not sure we need nibble separation rather
>than just two 16 bit fields as originally proposed.
>
>I think AS 45.1234 (equivalent to 2,950,354 or 45*65536+1234) is fine.
>
>I suspect that is what you meant, but, a nibble is 4 bits or one hex digit.
>
>Owen
>
>
>--On December 13, 2005 5:14:05 PM +1100 Geoff Huston <gih at apnic.net> wrote:
>
>>Your point about syntactical confusion with AS number communities is well
>>taken.
>>
>>What about a dot as the nibble separator within the AS number then?
>>
>>Geoff
>>
>>
>>
>>At 05:07 PM 13/12/2005, Owen DeLong wrote:
>>>In general, I like the policy, however, I think the nomenclature is a bad
>>>idea as specified below:
>>>
>>>>          Nomenclature
>>>>
>>>>          It is proposed to identify 4-byte AS Numbers using a syntax of
>>>>          <high order 16 bit value in decimal>:<low order 16 bit value
>>>>          in decimal>. Accordingly, a 4-byte AS number of value 65546
>>>>          (decimal) would be identified as "1:10".
>>>That fits really well with the current ASN:Comm syntax used for
>>>communities... NOT!  This will create no end of confusion by
>>>using this nomenclature.  Is 45:1234 AS45 community 1234 or
>>>is it AS45:1234?  What about communities within 16 bit ASNs?
>>>Will it then become AS45:1234:5678?
>>>
>>>Perhaps using dot as a separator as in 45.1234 instead would be a better
>>>idea.  Perhaps hyphen, perhaps something else.  I really don't care, but,
>>>I'd hate to see : used as it will do nothing but create confusion.
>>>
>>>Owen
>>>
>>>--
>>>If this message was not signed with gpg key 0FE2AA3D, it's probably
>>>a forgery.
>>>
>>>
>>>
>>>_______________________________________________
>>>PPML mailing list
>>>PPML at arin.net
>>>http://lists.arin.net/mailman/listinfo/ppml
>
>
>
>
>





More information about the ARIN-PPML mailing list