I store large models in the database because I need to have a historical data to compare to. That said, I could probably also automate a git repo but it will be just that much more work and git with binary files really doesn't make sense. Storage is really cheap and I assume the database stores bytesa types like they do text within a separate and partitioned section of the database.
Thanks,
~Ben
On Wed, Mar 13, 2019 at 1:37 PM Jamesie Pic <jpic@yourlabs.org> wrote:
Make dump/restore of database data unnecessarily expensive in terms of time and space imho.
