Hi all,
Since we're adding our nickles to this discussion ....
The debate centers on two different methods for extending a "standard". The
two methods both achieve the same end. One method (the use of multi-dimensional
coordinate variables) is clean and elegant, but is in some sense not backwards
compatible. The other ("referential attributes") patches the old standard in a
backwards compatible way but is not quite as clean.
NetCDF users are a broad community brought together by the need for standards.
This discussion has come together because of the need to agree upon a standard.
So which of these two methods will strengthen the standard? Which will increase
the overall usefullness of NetCDF -- not just for users of the particular
feature under discussion?
My bias is in favor of the change which is backwards compatible. That would be
my personal vote, if I were asked to cast a ballot on this issue.
Guess what, though? Noone is asking us to cast a ballot. I would argue that
which choice is taken is far less important than that **some** choice is taken
AND DOCUMENTED. Much of this discussion springs from the COARDS profile, a
written document that after many many hours of work documented an initial
(incomplete) set of decisions that had been discussed and resolved. That
document has proven to be widely valuable and has been a real force to promote
standardization. Who will volunteer to extend the COARDS document and lead the
discussion that arises out of it? (I'm afraid that this time around schedules
force me to rule myself out.)
- steve
--
| NOAA/PMEL | ph. (206) 526-6080
Steve Hankin | 7600 Sand Point Way NE | FAX (206) 526-6744
| Seattle, WA 98115-0070 | hankin@xxxxxxxxxxxxx
