CCNx status and relation to public-webpayments

Greetings,

I have watched the videos of the CCNx/NDN community meeting on Sept 9 
2011 (Program: http://www.ccnx.org/ccnxcon2011program/) and want to 
share my impressions here. I believe that CCNx development will be of 
interest and possibly important to this web-payments group.

Caveat: I don't program in CCNx; many of the presenters were 
foreign-language speakers; and the acoustics in the question periods 
were sometimes poor, not everyone waited for a mike, and no transcript 
has appeared. Together these things mean I can't always be sure of 
what was going on.

But, regardless, several things were so clear that even I couldn't 
miss them. :-) .

First, this is a large and active endeavor that crosses academic, 
corporate, and national boundaries: many universities were represented 
and have active development (Cambridge; UCLA; U. of Beijing; Colorado 
State; U of Bern; many more); and many large and small corporations 
and companies as well (the larger including Alcatel-Lucent-Bell; 
Toyota; Samsung; Intel). There also appear to be people from many 
countries involved; not only by the ethnic and language mix but by 
mentions of programs in other countries and universities (France, 
Switzerland, China, Korea, England, Japan).

Second, CCNx, which this community is coding, is only one aspect of 
the effort; there are other ways of expressing the concept and other 
names and centers. NDN – Named Data Networking – is another major way 
(http://www.named-data.net/); I believe that one is more directly 
academic-based; whereas CCNx, having been started at PARC/Xerox, has a 
more corporate emphasis – although I believe it's all open source in 
both cases. There are others. In any case some key people appear to be 
working in/on more than one of them, so they are closely related.

Third, more directly related to web-payments and to the main point I 
want to develop here, is the fact that almost all the formal 
presentations were about transport-layer configuration and efficiency 
problems, without reference to the overall purpose of CCNx/NDN and how 
it might benefit society as a whole.

This was addressed directly in the question periods, where the 
moderator asked for, and got, suggestions about things that were 
lacking in the overall roadmap for CCNx/NDN. One questioner, a senior 
academic (identified as "Karen"), asked extensive questions and made 
suggestions that included this critical core: (paraphrased from my 
notes): "Replacing the internet as we know it is a huge uphill job and 
we need to have a truly compelling set of reasons for it, or else why 
do it?" In other words, she was asking "What is the benefit to society?"

No general answer was given that I saw in the meeting; and my belief 
is that the work the web-payments group is doing *might* be the most 
important answer to that question (although there are others). Hence 
the need to remain aware of what both groups are doing.

In other words, as my thinking goes, if web-payments can become viable 
using current HTML, then CCN might not be necessary, and may not end 
up replacing the current Internet protocols. If, however, it turns out 
that one of Van Jacobson's predictions is correct: that the 
complexity, security, and distribution-bandwidth problems of 
point-to-point TCP/IP *preclude* the successful efficient initiation 
of web-payments by small players, then it may be that CCNx/NDN and 
web-payments need each other, necessarily. That is: Web payments may 
need CCNx/NDN in order to to be implementable for small players; and 
CCNx/NDN may need web-payments as a goal in order to get the impetus 
to overcome the huge inertia that is the entrenched HTML4/5 protocol.

If, after experimentation by web-payments, that *does* prove to be the 
case then at some point the web-payments group will need to port their 
work on top of a growing underlayer of CCNx/NDN transport protocol 
code. And for that to occur, there will need to be recognition and 
co-ordination in both groups.

In addition, as part of the question-periods, several other concepts 
were brought up and noted by the moderator as key lacks of the 
CCNx/NDN efforts to date: among these were, interestingly, business 
models, meta-data integration, and privacy issues. These have been 
worked on for years by ODRL and RDFa and other groups, and are being 
worked on (and in some cases co-ordinated) by web-payments. This makes 
it seem even more likely and beneficial that such co-ordination should 
occur between the CCNx/NDN and web-payments groups (if it turns out 
that CCNx/NDN is truly necessary).

Understandably, unless CCNx/NDN gets to a certain point of 
functionality, no-one developing in TCP/IP will move to it even for 
testing. But I think it's worth being aware that this might occur 
soon, and be ready. CCNx/NDN is already in testbed mode with tunnels 
though the internet linking PARC and several university hubs 
(http://www.named-data.net/testbed.html). In this regard I'm not the 
right person to know what's operational and what's not, although I did 
notice the interesting development of the javascript CCNx browser node 
(https://github.com/truedat101/ccn4b#readme; see also the .zip of a 
Poster explaining it, reachable from the CCNx program page linked 
earlier; presenter is Kordsmeier).

Given the CCNx/NDN people's questions about their own high-level 
purposes, I believe it would also be good if they could get 
information relating to the web-payments effort; in particular, I 
believe the CCNx/NDN people need a way to hear answers that come from 
people outside their specialized community. I sent them a short 
submission relating to this on their website feedback form, in which I 
included some ideas and links, including to the web-payments group 
pages. Of course, hopefully, someone from CCNx/NDN is already 
monitoring web-payments/ODRL/RDFa (etc.) development, and realizing 
that they might benefit from interfacing with it. If not, I hope that 
they will begin.


Steven Rowat

p.s. A final note about CCNx/NDN overall development: NDN is one of 
only four groups who are recipients of National Science Foundation 
grants for designing the future internet (Future Internet Architecture 
Awards); and this was narrowed down from 50 groups in an earlier 
choice process that took 5 years. (http://www.nets-fia.net/.) My 
reading of the four groups' summaries is that at least the first three 
(CCN/NDN; MobilityFirst; and Nebula) and possibly also the fourth 
(eXpressive) can probably co-exist and provide mutually supportive 
levels of architecture. In other words, perhaps by design, the 
narrowing-down process of the NSF awards may no longer be looking 
through a group of competing processes that will yield eventual 
winners and losers. They may now be funding things that will all come 
to pass, together.

Received on Friday, 14 October 2011 15:07:31 UTC