[dtn] Aside: the kind of charter I'd like

Stephen Farrell <stephen.farrell@cs.tcd.ie> Mon, 08 September 2014 10:29 UTC

Message-ID: <540D84F4.3030007@cs.tcd.ie>
Date: Mon, 08 Sep 2014 11:29:08 +0100
From: Stephen Farrell <stephen.farrell@cs.tcd.ie>
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:31.0) Gecko/20100101 Thunderbird/31.0
MIME-Version: 1.0
To: "dtn@ietf.org" <dtn@ietf.org>
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: 7bit
Archived-At: http://mailarchive.ietf.org/arch/msg/dtn/iIvX-Ei3LGx3L2Ue37KSuAqDnZA
Subject: [dtn] Aside: the kind of charter I'd like
Precedence: list

So as I said before a few times, I do think I'm in the rough
on this, but the text below is where I'd like to have started
in a charter discussion. At the BoF however, I thought that
I was less in the rough than I expected, so I'll send this one
mail and see if it gets a bunch of support. I don't expect it
will, in which case I'll stop asking:-)

In other words, if you don't like this, and prefer Marc's you
don't need to say so unless other folks first do say they like
this. That is, silence here == forget about it:-)

And btw, this is not something suited for the DTNRG. This
is pure protocol engineering and involves no research so
researchers will quite properly not spend time on this, just
as they have not spent time on any 5050bis effort.

Cheers,
S.


The DTN architecture (RFC 4838) and protocols developed according
to that architecture (RFCs 5050, 5326) have been successfully used
in small scale trials and by the space community. As a result a
number of relatively minor functional deficiencies have been
discovered (e.g. the need for relative timing for bundle expiry). A
number of those are documented in <draft-foo>.

However, despite this small-scale success, the bundle protocol (RFC
5050) has not seen widespread deployment and is not considered a
tool of choice for developers dealing with delay and disruption on
the Internet today.  Among the reasons for this are that it is not
easy to build a DTN from commonly used developer tools and still
interoperate with an RFC 5050 DTN. The fact that an RFC 5050 DTN
can't easily make use of deployed infrastructure (be that for
messaging or the web) also imposes a very high hurdle for
deployments.

The goal of the DTNWG is to develop a protocol that 1) adheres to
the DTN architecture, 2) that allows for relatively simple
gatewaying to an RFC5050 DTN, but that 3) is far more likely to be
deployed by developers using commonly available tooling and
existing deployed infrastructure.

The DTNWG has one deliverable (possibly in multiple RFCs) which is
a replacement protocol for RFC5050 meeting the above high level
requirements and the DTN functional requirements aleady met by
RFC5050 and co.

[dtn] Aside: the kind of charter I'd like Stephen Farrell
Re: [dtn] Aside: the kind of charter I'd like John Dowdell
Re: [dtn] Aside: the kind of charter I'd like l.wood
Re: [dtn] Aside: the kind of charter I'd like Ivancic, William D. (GRC-LCA0)
Re: [dtn] Aside: the kind of charter I'd like Burleigh, Scott C (312G)
Re: [dtn] Aside: the kind of charter I'd like Templin, Fred L
Re: [dtn] Aside: the kind of charter I'd like Templin, Fred L
Re: [dtn] Aside: the kind of charter I'd like Templin, Fred L
Re: [dtn] Aside: the kind of charter I'd like Stephen Farrell
Re: [dtn] Aside: the kind of charter I'd like Burleigh, Scott C (312G)
Re: [dtn] Aside: the kind of charter I'd like Stephen Farrell
Re: [dtn] Aside: the kind of charter I'd like l.wood
Re: [dtn] Aside: the kind of charter I'd like l.wood
Re: [dtn] Aside: the kind of charter I'd like Templin, Fred L
Re: [dtn] Aside: the kind of charter I'd like Marc Blanchet
Re: [dtn] Aside: the kind of charter I'd like Burleigh, Scott C (312G)
Re: [dtn] Aside: the kind of charter I'd like Stephen Farrell
Re: [dtn] Aside: the kind of charter I'd like Taylor, Rick
Re: [dtn] Aside: the kind of charter I'd like Sebastian Schildt
Re: [dtn] Aside: the kind of charter I'd like Stephen Farrell
Re: [dtn] Aside: the kind of charter I'd like Templin, Fred L