There are hard provenance questions which are based on the trustworthyness of provenance data - why should I believe this record that something was done. In general these require a trusted 3rd party provenance service, rather than scientists storing provenance in their own mIR. The myGrid project does not have the resources to tackle these issues ad will concentrate on softer provenance problems, e.g. what views over a provenance data set provide added value for scientists. Luc has one project http://www.pasoa.org that is addressing hard provenance questions, and has a further EU project proposal in preparation.