Bruce Lilly writes:
As I have always observed and practiced use of the term, "consensus" means:
+ there is at least one proponent of the proposition in question
+ there is no opposition on substantive grounds to the proposition.
The IETF instead uses ``rough consensus,'' which is defined as ``any
situation that the working group chairs think that they can get away
with declaring to be rough consensus.''
For example, the DNSEXT chairs sent the ``axfr-clarify'' document to
the IESG, even though seven people have publicly objected to that
document on various substantive grounds.
As another example, the IDN chairs declared ``rough consensus'' on IDNA
even though more than _three hundred_ people publicly objected to IDNA,
specifically to its failure to prohibit multiple names with identical
semantics. (There were many other objections, but IDNA's mishandling of
semantic similarity is what has really angered users.)
---D. J. Bernstein, Associate Professor, Department of Mathematics,
Statistics, and Computer Science, University of Illinois at Chicago