The process looks fine, if a bit overspecified. I have some concern about the choice of ombudsman*, given the extremely poor experience I had with ICANN's ombudsman a few years back, although the lack of pay will certainly discourage careerism. My larger concern is that you need at least some general descriptions of what would constitute harassment to keep it from being too arbitrary. It seems to me that the essence of harassment is unwanted attention directed to an individual or small group of individuals. If the target told the harasser to stop and he continued, that makes the claim more credible, although there are surely situations where the situation is threatening enough that it's reasonable not to respond at the time. A few general guidelines like that would make the policy a lot more concrete, and avoid "how could I know X would take that as harassment?" issues. R's, John * - ombudsperson is unacceptably species-ist, while ombudsman is a Swedish legal term from the 1950s