i haven't yet studied the new file formats, but i would
make the following general comment: it is preferable to
represent all data exactly once. including multiple
representations of the same data simply increases the
chance of discrepancies -- i.e., the chance of error.
seth
Patrick Nichols wrote:
Just an open question -- I've noticed that in the files I get from the
pipeline, portals are represented twice -- there is a portal list for
each space (by unique id) and then a separate set of portal files.
The question is: is there any real value in keeping the first set of
portal lists (per space)? in my code, spaces are made, THEN portals are
made -- so the only values seems to be in computing some space adjacency
stuff.
Thoughts?