Hi John,
It is true that there aren't that many clients that can parse the Dapper
conventions for in-situ data. But, as far as I know, there aren't that
many clients that can read *any* of the network protocols for
representing in-situ data (such as some of the IOOS implementations
using OGC standards). There is a chicken-and-egg problem here -- no one
wants to waste time developing client software for conventions that
won't be widely used, but conventions won't be widely used until client
software is available to read the data that conform to the conventions.
Someone needs to be willing to take the first step and take a chance on
developing client software for in-situ data that is compliant with some
kind of convention, whether that is Dapper or some other protocol. I'd
make the argument that, as a developer of libraries for parsing
different formats, that someone should be Unidata. The first step might
be to get a group of people together to evaluate the currently available
conventions (in terms of applicability, scalability, etc.) and come to
an agreement on which convention would be the best to use for an initial
implementation of an in-situ client library.
Cheers, Joe
John Caron wrote:
Rich Signell wrote:
NJ folks,
In the Common Data Model Scientific Datatypes docs, there is mention
of a "profile feature collection"
http://www.unidata.ucar.edu/software/netcdf-java/CDM/
http://www.unidata.ucar.edu/software/netcdf-java/reference/FeatureDatasets/PointFeatures.html
Profile data seem like a natural fit for OpenDAP sequences as a
delivery mechanism over the web and NetCDF4 as a persistence format,
so I'm wondering:
Is the profile scientific feature type designed to be compatible with
Dapper sequences for profile data?
Is there an example of a NetCDF4 file that contains a profile feature
collection?
Thanks,
Rich
Hi Rich:
We are working on netcdf3-CF encodings of all the Point Feature types:
https://cf-pcmdi.llnl.gov/trac/wiki/PointObservationConventions
Ill start to make a collection of sample files as I get them. We have
not yet started to look at ways to use the new features of netcdf-4
for this.
The new CDM Point Feature APIs all map well to opendap sequences and
dapper, although the actual work has not been done. One issue is that
I dont know of any clients who could make use of this. Do you have any
in mind?
_______________________________________________
netcdf-java mailing list
netcdf-java@xxxxxxxxxxxxxxxx
For list information or to unsubscribe, visit:
http://www.unidata.ucar.edu/mailing_lists/
