spatial reference

[Peter McCartney]

Ive been growing very unhappy with how much editing and documenting we are
doing in eml-spatial reference when it is really a wholesale import of the
esri profile of fgdc with no content change or additions. Ive checked into
an eml-mccartney branch three files that show an alternate way to handle
this that follows the stmml solution. basically, i redefined
spatialReference to be merely a choice between two elements - a name that
identifies a pre-defined projection or an element that is based on the spref
complex type from esriprof80.xsd (the esri profile of fgdc). I also added
eml-spatialReferenceDictionary.xml which would contained predefined elements
for all commong projections whoses names (projcn) appears in the enumeration
list for standardDefinition. 


This is not an entirely clean solution, as i cant right now validate
eml-spatialReferenceDictionary against esriprof80.xsd because spref is not
exposed as a root node. so it would take some fiddling, but its complete
enough to get the idea across.

the advantage is 1) we dont have to worry about docs or all this editing to
make it look like other eml modules, and 2) we can provide a list for users
who can recognize projections by name, but not know their parameters. This
follows a model similar to how ESRI's apps have users set projection
information.

Peter McCartney (peter.mccartney at asu.edu)
Center for Environmental Studies
Arizona State University
480-965-6791 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mercury.nceas.ucsb.edu/ecoinformatics/pipermail/eml-dev/attachments/20020925/abf6577e/attachment.htm