[clue] Publishing CLUE protocol as Experimental?

[Adam Roach <[e-mail]>]

CLUE working group and document authors:

During IESG evaluation, Benjamin and Ben raised some issues around the 
state machine design, timeouts, and interaction with the various SIP 
state machines. While these could probably be addressed by adding 
substantial new text and a handful of new states and supervisory timers 
to the CLUE state machines, as well as including additional detail about 
the interaction between such state machines and the SIP state machines, 
it's not clear to me that the CLUE working group has the kind of energy 
necessary for that kind of refinement at this point in its lifecycle.

An alternate path that both Ben and Benjamin would be okay with is 
publication of the protocol mechanism as Experimental rather than 
Standards Track, in order to allow the collection of operational 
experience around implementing and deploying the protocol with the state 
machines as specified. This would be sufficient to satisfy their 
concerns around the currently-defined state machine and would allow 
publication of the CLUE documents. (Note that, due to the relationship 
between clue-protocol, clue-signaling, and clue-datachannel, this would 
also result in the publication of clue-signaling and clue-datachannel as 
experimental.)

Recognizing that many people are out of the office during this time of 
year, I'd like to ask that anyone who has comments regarding the 
proposal to move the core CLUE documents to Experimental status respond 
before Friday, January 18th. Thanks.

/a