--On Wednesday, February 19, 2014 13:56 -0500 Barry Leiba
<barryleiba(_at_)computer(_dot_)org> wrote:
Barry, can you add a little more clarity on this point,
please? Does your decision mean that we shall not ever
normatively reference CBOR in future protocols?
It does not mean that. CBOR does remain as a Proposed
Standard. This is about process, and the key point is here:
Barry,
Purely as a process clarification (I don't care enough one way
or the other, but I haven't been involved in JSON work), is it
correct that, if someone wanted the state that Joe asked about
(i.e., no normative references to CBOR), they would need to take
one of the following three steps:
(1) Appeal your just-announced decision, asking that the
document be reclassified to Experimental, Informational, or, I
suppose, Historic?
(2) Generate an I-D specifying one of the above four results and
the reasons for it and try to get that I-D approved.
(3) Generate an I-D for an Applicability Statement that would
identify CBOR as "not recommended" and try to get it approved.
While that would not prevent a normative reference requiring
CBOR use, it would certainly touch off an interesting discussion
should anyone try to do it.
In either the second or third cases, the parties involved could
either try to get the I-D through the process as an individual
submission or could try to get some relevant WG, presumably (but
not required to be) JSON, to take it into their work program[me].
Does that correctly summarize the process alternatives?
john