On Mon, Sep 01, 2014 at 09:15:50PM +0200, Eliot Lear wrote: > On 9/1/14, 8:38 PM, Dave Cridland wrote: > > So we'll only discuss ideas that are written up as drafts, and we'll > > only discuss them in a working group, and we'll only create a working > > group around drafts that have traction, and we'll only consider a > > draft to have traction if there's discussion, and we'll only discuss > > them in a working group, and ... > > Only is a bit much, but the point was that a number of important > documents DIDN'T use the WG process, and it turned into a mess. I recall that BTNS WG was as difficult to get spun up as opportunistic security (OS) was to get in front of the IESG w/o a WG. The benefit of a WG is that once you have a charter the debate on that ends. But for something like OS the result would undoubtedly have been: more wasted energy, more time to publish. (It'd have taken much longer to get to published state. First we'd have needed a BoF. Or two. And charter bashing. We'd have to have found a WG chair(s). We'd have to have had the exact same discussions we already had. If at the end of WGLC people in the rough were passionate enough we'd have had more of the same on ietf@xxxxxxxx. What would have been the benefit? A WG would have to have had a more ambitious remit to be worthwhile. Such as applying OS to a variety of application protocols... that are currently the remit of other existing WGs, thus necessitating a lot more negotiation. If the issue here is that not using a WG means cluttering ietf@xxxxxxxx, well, we could separate IETF LCs from other topics. But really, we still need IETF LCs, and there's often going to be a lot of discussion. Nico --