Hi all,
Recent discussions on the wcsplus mailing list have included possible
objectives for the initiative. I'd like to make it clear that my view
is that an overarching goal is for the results of the work to feed
into the WCS Standards Working Group and to have that input based on
experimentation with actual implementations.
The adopted approach is to start with WCS 1.0 because there are
already a substantial number of working implementation (e.g. the
GALEON implementations) and because it appears that the next WCS spec
will differ significantly from WCS 1.1. Hopefully we can act quickly
enough that the results of the work can inform the WCS 1.2
specification, which seems to be moving in the direction of a "base"
)or "core") plus "extensions" definition of WCS.
In various discussions, the following list of specific objectives for
WCS 1.0 Plus have been put forth. In keeping with the WCS 1.0 Plus
philosophy,
these should be minimal changes that are relatively easy to implement and test.
1) "Clean up" the WCS 1.0 specification making it fully based on the ISO
19123 coverage specification
2) Sort out some of the inconsistencies between the KVP and XML
schema documents.
3) Extend the WCS 1.0 specification adding the following features:
a) change the range set model in order to support multi-
parameters -- in agreement with the ISO 19123 model.
b) clean up and extend the CRS encoding --in agreement
with the ISO 19123 model (including adoption of 07-112
proposal for 19123 irregular grids in GML).
c) minimally 're-interpret' 1.0 AxisDescription for
heterogeneous coverages.
d) add negative Z dimensions
4) Extend the "blessed" format list by including the CF-netCDF
5) Come up with an encoding specification (at least at the
implementation level) for the CF-netCDF datasets returned
by a getCoverage request.
a) binary
b) XML document + binary ??
6) Adopt a WPS approach for asynch, proven in the DEWS WCS. (Note that
there is some concern that this may be too ambitious for rapid
implementation in the initial phase of WCS 1.0 Plus)
I am sending a copy of this to the GALEON list because a key question
is whether GALEON participants think this is a more productive
approach than attempting further implementation and testnig of WCS 1.1
at this point.
-- Ben
