[EMAIL PROTECTED] wrote on 08/09/2006 06:37:25
AM:

> Hey all, perhaps especially Martin,
>
>
> In looking at the code of the interface
>
> http://geoapi.sourceforge.
> net/snapshot/javadoc/org/opengis/referencing/cs/SphericalCS.html
> and perhaps this is the current form:
> http://svn.sourceforge.
>
net/viewvc/geoapi/trunk/geoapi/src/main/java/org/opengis/referencing/cs/SphericalCS.

> java?view=markup
>
> It's just a shell. There is nothing there preventing the interface from
> being an 'degenerate ellipsoidal' crs.
>
> 1) How does this shell imply that implementations shall be classes like
> geotools' DefaultSphericalCS? Is this simply a request that implementors
> not create derivatives with the distance axis a height_above_ellipsoid?
> (This question is about how geoapi works not about what a sphericalCS is
> supposed to be---I get what is implied mathematically but not how geoapi
> expresses this expectations of others.)

ISO 19111 says this is a 3D coordinate system (always).  {Radius, angle,
angle}  A 3D degenerate ellipsoidal CS is {ellipsoidal height, angle,
angle}.  The average error you would get by confusing
GeocentricCRS/SphericalCS with GeographicCRS/EllipsoidalCS is 6000km (give
or take). :) Zero does not mean sea level. :)

>
> 2) Could this interface be implemented by a PolarSphericalCS? That is
> substituting a \Phi coordinate (angle from the rotation axis) for the
> \Psi coordinate (angle away from equator). Would such an implementation
> go against the intent of the interface? We either need to note the
> distinction in both the geoapi javadoc and the geotools javadoc, or we
> need to change the geotools javadoc to state that it has chosen to
> implement *not* a polarSpherical but a ??Spherical coordinate system.

Yes, by reason of that's what it's supposed to be. :) This can be used in
conjunction with a GeocentricCRS.  Want to model radar or do
photogrammetry?  Use an EngineeringCRS with a SphericalCS, in which case,
\phi is angle from zenith or nadir.

In general, the CS interfaces _are_ shells.  Implementations (with code)
must ensure that the UsesAxes associations obey the constraints of the
standard.  This is too convoluted to enforce with interfaces, I think.

Bryce


-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
_______________________________________________
Geotools-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/geotools-devel

Reply via email to