[e2e] Transport Protocol Road Map(s)

Bob Braden braden at ISI.EDU
Fri Sep 5 14:06:46 PDT 2003


There are a lot of RFCs today.  Those of us who have been around awhile
have at least a general notion of the order and importance of various
technical developments in the Internet protcol suite.  Yet, when someone
asks me which RFCs currently rule in the SMTP space, for example, I am
pretty much at sea.  There are a lot of newbies in our midst these days.
We expect them to digest the corpus of relevant RFCs to come up to speed,
but we don't give them much guidance in this task.

It has been suggested that we need per-area roadmap RFCs, which provide
snapshots of what is important (and briefly why) in a particular subject.
These of course would rapidly become partial truths, but it is easier to
update than to write originally, and a partial truth is better than no
truth.

If you search for "TCP" in the RFC search engine, you get 119 hits.
Maybe 3/4 of these are obsolete or irrelevant, but that still leaves a
passle of documents.

This message is a challenge to the end-to-end community: we need a
serious effort to draft roadmap(s) for the transport area.  How many
RFCs does a TCP implementor today need to have at hand?  Primary?
Secondary?  And simularly for UDP, RTP, and SCTP.  Does it makes sense
to have a combined roadmap for the transport area, or should there be
separate ones for each transport protocol?

Step right up, please.

Bob Braden




More information about the end2end-interest mailing list