The other Scott's approach looks like it's clearly the most reasonable,
and follows a model we have used before. No reimbursement for
performance of services; no reimbursement for meetings that are
associated with IETF; reimbursement for travel to special (not
IETF-associated) meetings where necessary. As usual, we have to be
careful not to get too specific. This might be just specific enough.
Ietf mailing list