Ignoring featureid

[Thomas Hirsch]

I am working with continuously updated datasets that are being imported from a source that does not use unique feature ids. Tools like ogr2ogr then assign feature ids at will upon import. 

If I import datasets from the resulting PostGIS database into a geogit repository, it seems to only look at featureids to identify changes. Is there a way to obtain the opposite behaviour and ignore feature ids? To identify my changes (or rather the common ground), I would want geogit to first look for identical geometries (identifying potential changes in the attributes), then -optionally- identifying overlap in other key attributes: name, etc.- one could probably look at everything that has an index.