[CF-metadata] Question on WKT representation of CRS

From: Jonathan Gregory <j.m.gregory>
Date: Mon, 3 Oct 2011 13:53:42 +0100

Dear Phil

It's a good thing you're working on this and I'll be interested to see the
proposal. I'd like to anticipate what concerns I might have (of course these
are just my views) based on other initiatives to support the use of other
conventions in CF, which is obviously a good idea if we can do it. I think
points to consider would be

* Is the metadata self-explanatory to a human reader, so that the dataset is
self-describing, as generally intended by CF?

* Is there any possibility of redundant information, and hence inconsistency,
between the external convention and CF metadata, or within the metadata of
the external convention? If so, can we write down rules so that inconsistency
could be automatically detected e.g. by the CF checker, and to advise a data
user how the metadata should be interpreted if there is an inconsistency?

* What is the process for making sure CF and the external convention continue
to remain compatible?

Best wishes

Jonathan
Received on Mon Oct 03 2011 - 06:53:42 BST