[Stackevo-discuss] PLUS side meeting in Seoul, Tuesday 12:15

[Brian Trammell <[e-mail]>]

Greetings, all,

The PLUS proponents will have a design side meeting in the Park Studio room on the 5th floor of the Conrad Seoul tomorrow, Tuesday 15 November, at 12:30. The purpose of this meeting will be to decide what would need to be done before a second attempt at a working group forming BoF, and to find volunteers to help working on  the drafts necessary. So, if you're interested in PLUS and want to help, please come by.

If you're coming to the meeting, please let us know, as space is quite limited in the room (about 20). Also, make sure you get something to eat beforehand (or on the way there), as there will be no lunch served.

If you're interested in helping but are not free at lunchtime tomorrow, please let us know that too.

There are three (new) related drafts, now:

(1) https://tools.ietf.org/html/draft-hardie-path-signals describes the type of signaling inferable from current transport protocols, which PLUS aims to replace in a transport-independent way; it's derived from Ted's talk at the PLUS BoF in Berlin.

(2) https://tools.ietf.org/html/draft-trammell-plus-abstract-mech is a more-complete description of the abstract mechanisms for signaling described in my slides in the PLUS BoF in Berlin.

(3) https://tools.ietf.org/html/draft-trammell-plus-statefulness defines a minimal transport-independent state machine for state-keeping middleboxes and a set of abstract signals for driving this state machine, to replace the path signaling in (1).


We do have a rough sketch of the way forward:

(1) We have tightened the scope of the proposal (see the current proposed charter at https://github.com/ietf-plus/charter/blob/master/charter-plus.txt), restricting it to replacing those path signals that are commonly assumed to be available for on-path state management (as discussed in draft-hardie-path-signals), implementing the mechanism described in draft-trammell-plus-statefulness.

(2) We would like to attempt to have another BoF, but not without having a protocol document, defining a protocol which solves the more restricted state signaling problem. The discussion will go much more smoothly with bits on the wire to argue about. This protocol will be largely based on the statefulness draft, above, while the abstract mechanisms in draft-trammell-plus-abstract-mech would be used for timeout discovery and exposure.

(3) We have requested a [e-mail] mailing list; this is in processing and should happen soon. We'll announce here when it does. We should have done this before the PLUS BoF in Berlin, actually. After this move, the SPUD list will be closed (as that experiment has completed); please subscribe to [e-mail] once it's up if you're interested in following the PLUS effort.

Cheers,

Brian and Mirja