Peter Gutmann wrote:
Sean Turner <turners(_at_)ieca(_dot_)com> writes:
Russ and Tim have challenged me to try to advance CMS to Draft Standard
using the process in BCP 97 to deal with the dependencies that have not
yet advanced (or aren't advancing). To that end, I've drawn up the
following questionnaire that I'd like implementers to complete and
return to me. What we're shooting for is to find two implementations
that generate messages for each feature, and two implementations that
process each feature. We don't think it should matter if the same two
implementations are used in both stages.
Are you just after straight yes-or-no answers, or are you interested in
metadata as well? By metadata I mean things like "I probably support method
XYZ but since I've never found anything else that does to test against I can't
guarantee that it's correct", or "I've implemented XYZ but since no user has
ever asked for it it's never been tested". It may be useful to get (or at
least I'd be very interested in seeing) data on what's actually being used in
the real world. Motivated by the history of RFC 4134, I'd also prefer to be
able to qualify a claim of "I support XYZ" with further details if there's
never been any opportunity to test whether it's actually implemented as
required (in other words to differentiate "I'm pretty sure I support XYZ" vs.
"I definitely support XYZ and have interoperated with others using it").
I'm definitely interested in metadata.