[icnrg] fixed header self-declaration (was Re: Fwd: New Version Notification for draft-lindgren-icnrg-efficientiot-02.txt

[[e-mail]]

On Thu, 8 Jan 2015, Mark Stapp wrote:

> hmm: I think this is a bit strong:
...
> I appreciate your advocacy for "named functions", but that capability seems 
> to me to be one of the various interesting services that an ICN-ish network 
> might support. I'm not convinced that the concept needs to be included in 
> every discussion or accomodated as a core property.

point taken, no problem. Suggesting that I ask for "accomodation as a 
core property" might be strong, too? ;-)

However, I consent with the extremist label for myself when it comes to 
multi-protocol support. For example, I would be very much in favor of 
self-declaration in form of an additional type field in the fixed header 
saying "I'm carrying ccnb stuff", "I'm cisco2015", "I'm NDN". This would 
clarify what the version field is about: it's the version of the fixed 
header and nothing more, not the version of the carried protocol data 
units.

Views?

best, christian


> I don't think this is "tyranny" or even "extreme". CCN/NDN has a couple of 
> core properties, and we're trying to discover whether those properties 
> compose a system that actually ... works. getting some focussed discussion on 
> what is actually in a name, for example, seems ... far from "extreme" to me. 
> it is not glamorous to be sure, to ask folks to say "we have to have 
> such-and-such available" or "we must NOT have such-and-such", and to ask them 
> to write that down, and support those assertions. but I think that's a good 
> way to highlight the underlying assumptions and semantic issues about what 
> NDN can express, and how.
>
> -- Mark