Standardizing reputation query mechanisms

Lately, despite the low level of activity on this list, there has been
a great deal of activity in the development of a new generation
reputation services beyond the existing blacklist/whitelist paradigm.
Several (including one I am working on) either are available in some
state currently or are in development This is an exciting development
for the email industry in general, but does present some challenges.
The primary one is that most of these services are commercial to some
extent or another which makes sharing information difficult.
Despite this, an overwhelming concern I have heard from ISPs and MTA
vendors is that each of the developing services is publishing its data
in a slightly different way, and beyond that several protocols have
been suggested as standards for querying this data.
An area where it should be possible for all of us to work together
without the problems of commercial damage is in development and
deployment of a standard protocol for publishing and querying
reputation data. This problem is much more complicated than it may on
its face appear, because reputation services will have a much wider
range of sematics than traditional blacklists. They will return
different levels of granularity of data, from a single binary score,
to a huge range of scores over individual data points or even
customized scores for individual queriers. Or tey may provide a list
of suggested actions to be taken. They may have a need to allow email
receivers to guide the semantics of the query (e.g. I want elements
x,y,and z but not the other 23).

The advantage of getting us all to agree on the mechanisms to access
and exchange this data is that the mechanism can be built into every
MTA (if desired) and all of the systems will be supported without need
to develop new libraries every time someone creates a system. A
standard protocol also makes it more straightforward to compare