Matt S Trout wrote: > > Thoughts on whether make_column_dirty should clear the deflated value if an > inflated one is present, guys? > So now that ::FS is fixed can we do this?