On Thu, Apr 7, 2016 at 6:09 PM, Brian E Carpenter <brian.e.carpenter@xxxxxxxxx> wrote:
On 08/04/2016 08:52, Michael Richardson wrote:
...
> > On the other hand, I realize now that previous thread went on longer
> > than I remembered, and there was a proposal that we (I, probably)
> > construct an RFC3933-style process experiment and let that run for a
> > while. If it works well, we can codify it by adding it to RFC7437bis.
> > So I'll do that. If anyone wants to volunteer to collaborate on it,
> > please contact me directly.
>
> Yes, let's do that!
>
> It would awesome if we could say definitely that the new rules
> contribute to more volunteers before we actually use them.
That's not quite how RFC3933 works. You'd *actually* run the experimental
procedure for (say) one cycle, with automatic reversion to RFC7437 unless
RFC7437bis was approved. I think it's a good idea. A one-year experiment
affects ~half the IESG and IAB seats so is highly unlikely to lead to
disaster.
Brian