ietf
[Top] [All Lists]

Re: IETF areas re-organisation steps

2014-12-26 14:17:10
On Fri, Dec 26, 2014 at 5:25 PM, Dave Crocker <dhc(_at_)dcrocker(_dot_)net> 
wrote:

On 12/26/2014 8:50 AM, Paul Hoffman wrote:
Maybe we can take a short hiatus from our relentless cute naming and
call it what it is: UPPER.

That's reasonable.

An alternative that is equally un-cute but highlights one of the more
salient aspects of this realm of work:

   End2End (E2E)

The technologies in this new area all reside in the end systems, rather
than the underlying packet-switching infrastructure.


But where do the endpoints lie?

The funny thing about the mail system we are using for this conversation is
that it is not end-2-end. The application endpoints are mail clients and/or
Web browsers. There are multiple transport endpoints, the inbound and
outbound MTAs, spam filters, the list serve...

When we get into the trust and security issues the endpoints are people and
organizations, not machines.


I see a division between two types of 'applications' work. One is work
defining actual applications. Things like Jabber, Mail, Minion Rush, Chat,
etc. In the early days of the Internet, the IETF had 100% of the
applications work. Today there are tens of thousands of applications in
widespread use. It is clearly beyond the capacity of the IETF to support
development of all such protocols even if it tried.

The other type of applications work is to provide standardized
infrastructure that can be used to build applications. So just like there
are standards for screw sizes, pitches, heads, etc. and these are vitally
important to people who make cars, boats and airplane engines, there should
be standards for applications infrastructure, like how to map an API to a
JSON based Web Service.