[LRUG] Field definitions in AR models

[Thomas Buckley-Houston]

Thanks Frederick Cheung. Your answer has made the most sense to me. I
can see why model fields were sacrificed now.

Just thinking out loud, but it wouldn't be the worst case of foregoing
DRY if fields were in the model *and* the DB. It reminds me of C's
header files, there's something polite and hospitable about having a
snapshot overview of what is 'hidden underneath'.