[ppml] Policy Proposal 2005-9: 4-Byte AS Number

Geoff Huston gih at apnic.net
Thu Dec 22 20:42:09 EST 2005

I took the lead from the name of the IETF draft ("draft-ietf-idr-as4bytes") 
rather than the document's title or its body (which uses "octet").

Frankly, I didn't think much about this choice at the time, and I've no 
particular opinion either way and if folk think that its more consistent to 
use the terminology "octet" in the place of "byte" in the proposal then I 
have no problem in making such a change.


  At 12:32 PM 23/12/2005, Scott Leibrand wrote:
>On 12/22/05 at 2:25pm -0800, Owen DeLong <owen at delong.com> wrote:
> > > I don't believe that "2-byte ASN" and "4-byte ASN" constitute obscure
> > > terminology.
> >
> > While I agree in principle, and, do not advocate rewording the policy
> > in terms of integer ranges, I do think it would be an improvement to
> > speak in terms of 16 bit and 32 bit.  After all, while the term byte
> > these days usually refers to an octet or 8 bits, this has not always
> > been true.  In fact, the term byte actually can mean anywhere between
> > 5 and 9 bits, depending on machine architecture, encoding scheme,
> > etc.  Byte has never been an unambiguous term.  This is one of the
> > reasons almost every RFC is written in terms of bits and octets
> > and use of the term byte is rare indeed.
>That's a good point.  The IETF work on this refers to 4-octet ASNs, not
>4-byte.  I would have no objections to a s/byte/octet/g replacement,
>though I'm curious if Geoff or anyone else has any comments on the
>PPML mailing list
>PPML at arin.net

More information about the ARIN-PPML mailing list