Not sure if y'all were familiar with this, but it might present some interesting challenges for table-based hcal and hcard data: http://actsofvolition.com/archives/2006/june/selecting Chris