Hi Jon,
As for INSPIRE, the European Directive asks Member States in article
11(1) (c) to establish and operate a network of "download services,
enabling copies of spatial data sets, or parts of such sets, to be
downloaded and, where practicable, accessed directly".
According to the last version of the INSPIRE Service Architecture, a
download service supports: download of a complete dataset or
datasets, or a part of a dataset or datasets, and where, practicable,
provides direct access to complete datasets or parts of datasets.
Thus, it is clear the need for "access" services as far as Feature
and Coverage-based datastes are concerned. As the INSPIRE directive
advises to utilize existing standards, "OGC service bindings are
taken as a guidance". Moreover, and more important for our discussion
about OPeNDAP, the draft IRs states that "taking all requirements,
opportunities and riscs into account, the default
communication-protocol and binding technology for INSPIRE services
should be SOAP (document/literal)". Therefore, OGC WFS, and WCS with
SOAP bindings seem to be a natural choice. Specifications maturity
will be considered, indeed.
However, the INSPIRE technical WGs are still analyzing whether the
SOAP based approach is feasible for the INSPIRE network services. In
fact, they recognize that the "usage of REST APIs and related
technologies may help open up INSPIRE resources to mass market
information infrastructures".
A possible solution might be to see the INSPIRE Network Services "as
a mediator between the services provided by the member states or
offered by third parties and their EU-level usage for example via the
INSPIRE geoportal.... Therefore services in the member states are not
required to be changed because of INSPIRE.
INSPIRE just requires that data and services in the member states are
made available through INSPIRE conformant network services". A
valuable case in point would be the OPeNDAP-WCS/CF-netCDF extension
specification, we have been discussing about.
It's certainly true that there's a feeling amongst many people that
OGC=interoperability and therefore any OGC-accepted solution is
politically correct.
ISO TC211/OGC deal with interoperability solutions for "Geo-spatial
Information" communities; this information domain is a
multi-disciplinary one, indeed. Thus, if we consider the FES
information domain as part of the geo-spatial domain, OGC/ISO
specifications should be considered as "general" interoperability
solutions to interconnect FES to the other Geo-spatial information
domains (e.g. GIS, Biodiversity/Ecosystems, etc.).
Questions are about how general they should be, and which "common"
semantic and structural level they should consider. The present
ongoing discussion on WCS/CF-netCDF Vs OPeNDAP has been providing
many valuable arguments to better understand Communities
interoperability. In fact, FES information domain is characterized by
well-accepted, specific and effective disciplinary specifications;
while, the GIS domain models and protocols are really close to the
OGC ones. Other interesting examples are represented by Earth
Observation and Biodiversity Communities.
Cheers,
--Stefano