Hi GALEON WCS collaborators,
If we are going to have a specific proposal in place for a netCDF encoding
profile for WCS by the time of the June OGC Technical Committee meeting, we
have to come to some agreement soon on what that proposed profile will be.
Currently, for the netCDF encoding profile, we have three draft documents which
serve somewhat different purposes.
1. A brief outline document with pointers to external information sources. The
main headings in
http://www.unidata.ucar.edu/projects/THREDDS/GALEON/netCDFprofile-short.htm
are seen to be the important ones.
2. In addtion, there is a need for at least one document that provides people
with overview information regarding the relationship between the OGC WCS
specification and the encoding profile format. In the case of netCDF, there is
a draft (by committee but a very good committee) of such a document at
http://www.unidata.ucar.edu/projects/THREDDS/GALEON/WCSnetCDFencoding.htm
This version is fairly informal and only a few pages long which was one of our
original goals.
3. A third more formally crafted, detailed delineation of the mapping between
the ISO data model and the CF-netCDF data model was written by Stefano Nativi
and Lorenzo Bigagli.
http://www.unidata.ucar.edu/projects/THREDDS/GALEON/Stefano-netCDF_profile_for_WCS_3.doc
Part of my concern is that, if we are going to replace the WCS list of 5
encoding formats with a requirement for document a proposed encoding format,
what we do for netCDF will set a precedent for WCS encoding format
documentation requirements.
To start the discussion, I'll say that my sense right now is that each of these
three documents is needed for a proposed new encoding format. It is important
to have pointers to information specific to the encoding format, another
document of 3-5 pages in length that provides an overview of the relationships
between the encoding format and WCS, and finally a detailed mapping between the
data models, complete with UML diagrams. Perhaps the latter two could be
combined into one.
But it will be crucial to get input from others -- especially from those
involved in trying to create WCS clients and servers that handle various
encodings. What's the minimum amount of information you need for the
implementation? What would be a more complete set of information? Are there
glaring omissions from the drafts we've put together for the netCDF profile
documentation? How can the documents be kept concise without omitting
important materials? The GALEON Phase 1 experience hopefully provided some
hands on experience for some of you trying to implement WCS for netCDF. Would
this information have gotten the job done more quickly. Some of you are
thinking about similar documents for other encoding format profiles. Will
these drafts be a reasonable template?
If we are to propose something formally at the June TC, we have to submit the
proposal 3 weeks in advance. That means we have to have the proposal finalized
by the beginning of June. I will be out of the office at conferences for most
of the rest of this month, but I should be in email contact most of the time.
I suggest we keep the galeon@xxxxxxxxxxxx.eduemail list involved in the
discussion because there still interested people in that forum who are not yet
members of OGC.
Have at it.
-- Ben
PS I was hoping to have this set up so we could carry on the discussion in the
context of the GALEON OGCnetwork, but I'm just not comfortable that we could
all spin up on that approach in time to get this done.