At the recent IETF meeting in Beijing there was a lot of activity that is of interest for SAIL.
In this post I will concentrate on those relating to Network of Information (NetInf). Overall it can be said that the issue of how to name information objects pops up in a lot of contexts. It is becoming apparent that for things like caching and streaming and p2p it is beneficial if you can name information objects directly. In this way it is easier to see if you already have a copy of the object. Below are some observations from work groups (WGs) and BOFs.
Namebased sockets BOF
Large interest including all relevant ADs. The BOF was chaired by Christian Vogt (Ericsson) and Martin Stiemerling (NEC Labs). The main draft was presented by Javier Ubillos (SICS). There were also presentations from Apple and Microsoft. It is clear that there is a strong need for internal coordination regarding the naming issue for the future internet.
IoT BarBOF
Fairly large confusion about what the scope should be (as can be expected at a BarBOF). Trying to figure out how to structure the work and which issues to address. It was pointed out that IRTF can hardly expect to be leading the work in this area as there are so many large projects addressing this going on in the research community. IRTF needs to find a niche or some type of coordinating role. It was suggested that it might be good to focus on some particular issue. Naming was one issue that was brought forward. Short papers on suggestions on scope and issues will be invited before the next IETF when they expect to have a regular BOF. We should make such a contribution bringing up ideas on object to object communication as well as one on the NetInf naming scheme.
DECADE
The work is progressing well in a good atmosphere. The chairs invited the presenters for an informal lunch meeting before the IETF session. The problem statement and survey draft are moving towards WG last call. We will contribute a section on NetInf to the survey document before that. There is now a first version of an architecture document. The requirements have been made more strict and contains now sharper and more focused requirements. Thoughts on how to make the design have been moved into the architecture document. Börje Ohlman (Ericsson) presented the draft draft-ohlman-decade-add-use-cases-reqs-02 which was very well received and those of the requirements that has not already been adopted into the requirements and architecture drafts already, since last meeting, will be done so after some further refinement. Especially the requirement on naming needs further word-smithing. DECADE is now ready for further discussion on naming. We should make a new version of the PPSP draft on naming which is focusing on the particulars of DECADE. A key issues to address is how to make it possible for existing P2P applications to use their existing naming schemes and still take advantage of DECADE. Another key issue is to make it possible for future applications to share the same object using a common DECADE name space.
PSPP
The Secure naming draft was presented by Ove Strandberg (NSN), though it only got 10 min at the end of the session it was well received. It can be noted that David Bryan went up to the microphone to recommend people to really read the draft as well as the NetInf naming paper that was published at the Global Internet Symposium in the beginning of this year. Next step for naming in PPSP is to formulate requirements for naming that can go into the requirements draft. We should write a draft for next IETF suggesting such requirements.
Disclosure and disclaimer: I am engaged in the SAIL project, on behalf of Ericsson. However the opinions expressed in this post are my personal, and not those of the SAIL project or my employer.
