On 11/10/2010 10:55 AM, James Gallagher wrote:
Hello,
My apologies if there is a better place to post questions about NCML...
I've been working on the automatic generations of aggregations using
NCML. That is, using software to crawl servers and build/write NCML
files that form aggregations of files on the crawled servers.
Question: Is there consensus on the best way to represent time
(assuming it is the new dimension) in a JoinNew aggregation?
Background: I have the code working and it's currently using ISO 8601
to represent time, although I can easily change that. This is work
that comes from crawling to build metadata records for dataset
catalogs - aggregations fit into this because they are an effective
way to mitigate some problems with very large metadata records choking
catalog software (and some other semantic issues regarding
'inventories').
James
--
James Gallagher
jgallagher at opendap.org
406.723.8663
_______________________________________________
thredds mailing list
thredds@xxxxxxxxxxxxxxxx
For list information or to unsubscribe, visit:
http://www.unidata.ucar.edu/mailing_lists/
Hi James:
Do you mean the best way to represent time in the "coordValue" attribute
on the netcdf element? That is the way I was doing it, and I trie to get
CF to accept those ISO strings as valid coordinates. I failed, so I
think I switched to using just numbers. So I think its reasonable to use
ISO dates, and I accept that in the CDM, but its not CF compliant
unfortunately.
John
