Re: [icnrg] Post IRTF Po0ll updates to icnping and icntraceroute posted

[Colin Perkins <[e-mail]>]

Hi Dave, all,

Thanks for the update. I agree these are ready to progress, and will 
start the IESG Conflict Review now.

Colin


On 10 Jul 2023, at 18:41, David R. Oran wrote:

> Hi, adding to the pre-IETF noise level, but anyway:
>
> I just posted updates to
> https://datatracker.ietf.org/doc/draft-irtf-icnrg-icnping/ and
> https://datatracker.ietf.org/doc/draft-irtf-icnrg-icntraceroute/
>
> which have passed IRSG poll. There were helpful comments from Brian 
> Trammell, Marie-Jose Montpetit, and Ari Keränen, and also a reference 
> correction from Hitoshi Asaeda.
>
> Therefore, the changes are:
> - fixed the reference to CCNInfo to point to the published RFC9344.
> - clarified the base64 encoding of nonces with the text suggested by 
> Ari.
>
> I believe these are ready to go to the RFCed now.
>
> I also replied separately to Brian’s comment, which I reproduce here 
> so you don’t have to go find it in the archive.
>
>     Brian: My only question is high-level, and of the form "why didn't 
> you write a
>     completely different document?" and as such can be safely ignored: 
> the design
>     (intentionally) sticks as closely as possible to an analog to 
> (Paris/Tokyo)
>     Traceroute, which has the advantage of least surprise for people 
> familiar with
>     IP networks. Have there been attempts to implement an analogue to 
> IP Record
>     Route for ICN?
>
> There’s not an exact analog, but path discovery in ICN Pathsteering 
> does record the path as an Interest traverses the net, with the result 
> visible to the producer, and then mirrors it back to the consumer in 
> the responding Data message. However, the recorded path is in the form 
> of identifiers local and meaningful only to each forwarder on the path 
> and hence are (intentionally) not directly usable by a third party as 
> the results of having done a traceroute would be. In fact there’s an 
> option to encrypt the path record to foil certain attacks.
>
> Now, if the purpose of doing the Record Route in IP is primarily to 
> use it to construct a source route for a later packet, then 
> Pathsteering accomplishes that as its primary raison d’être.
>
>
> DaveO