That is why I think this is an area that the IAB should look at.
I don't think we should be worried about running out of protocol numbers.
Deployment of IPv6 and multicast pretty much demonstrate that it is a non-issue.
At the application level we just need to persuade people that, no it is not
acceptable or sustainable to have an internet architecture that is dependent on
parsimonious assignment of IP Port or DNS RR numbers.
Where we have a resource code consumption issue we should either declare the
protocol essentially closed for extensions or work out a way to introduce a
sustainable extension mechanism.
So for example on DHCP I think it would be entirely justifiable to say 'this is
an infrastructure for assigning IP addresses and specifying the domain name for
the LAN and we do not therefore anticipate assignment of additional codes on an
ongoing basis'. This does not rule out special pleading (e.g. geopriv) but
clearly signals to people to think somewhere else.
Alternatively if we think that every new protocol is going to need a DHCP slot
to advertise its existence then we would have to work out a way to insert an
extensible mechanism.
In the case of DHCP the first approach is clearly the way to go.
-----Original Message-----
From: Jari Arkko [mailto:jari(_dot_)arkko(_at_)piuha(_dot_)net]
Sent: Wednesday, June 13, 2007 10:38 AM
To: Hallam-Baker, Phillip
Cc: ietf(_at_)ietf(_dot_)org; iesg(_at_)ietf(_dot_)org
Subject: Re: IANA registration constraints
Phillip,
My personal view is that we should develop an Internet
architecture that allows an infinite number of new protocols
to be deployed without consumption of scarce resources, i.e.
port numbers of DNS RRs.
...
So in summary, the IAB should be charged with identifying
the set of finite resources that IANA assigns and propose an
Internet architecture in which deployment of new application
layer protocols does not cause any of the finite resources to
be depleted.
I'm definitely in favor of improving the situation. And for
applications protocols this is probably an easier problem to
begin with. And as I said in the previous e-mail, as far as I
know, most new work uses field sizes and types that have less
scarcity.
However, the Internet runs to a large extent on protocols
that were designed decades ago, and some of those protocols
have number spaces that are very finite. I don't want to run
out of protocol numbers, DHCP message types, etc.
Jari
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/ietf