I started with a small stack of papers that a friend of the family had assembled some twenty years prior, and from then on it was digging through online archives. I’m interested in provenance so being able to keep citations is a major asset for me.
A second major drawback for me is that Gramps’ own file formats are not quite as stable and consistent as one would hope, and DTDs are harder to read than XML schemas. However I’m hoping to avoid all that by refactoring Betty’s current parsing of Gramps XML into integrating with Gramps’ Python APIs directly.