cnreediii
commented
4 years ago @UnclePoole - Yes, this is a known issue. As editor, I started the transition to terminology as used in the OGC and ISO. Unfortunately, we did (do) not have access to some of the original graphics and artwork so we are "stuck" with the original terminology in those graphics and tables. Also, we had to figure out how radical the terminology change should be. I forget the actual discussion but perhaps we should have gone further in the transition to OGC/ISO terminology. We could do that easily enough - especially given that the first version of Simple Features drew from the model used by Shapefiles! This is why we used "polygon" and not "surface". Also, as you point out, "surface" is a higher level abstraction. AS for Lineal, we probably should have transitioned to "linestring" and been more consistent.
The vector geometry terminology currently used in CDB is not consistent with the geometry types used in GGDM nor with the OGC Simple Features specification. The terms also do not seem to be consistent within different parts of the CDB 1.1 specification.
The terms in the feature dictionary are "Point", "Lineal", and "Areal" in the "Recommended Dataset Component" element of feature definitions. These terms are also widely used in descriptions in the OGC CDB 1.1 specification including Table 5-27 that describes which attributes should be present in which datasets. Various footnotes (19, 70) state that these terms have been replaced by "Point", "Linear", and "Polygon" and should be considered equivalent; however this does not appear to have actually been consistently done in the specification - different sections mix and match these various terms.
The GGDM terminology is "Point", "Curve", and "Surface" with an additional type of "Table" to represent non-spatial features describing metadata or informational characteristics. These align with the equivalent OGC Simple Features terms 6.1.4 Point, 6.1.6 Curve, 6.1.10 Surface. However, note that Curve and Surface are abstract types in Simple Features and the actual CDB implementation would more closely match 6.1.7 LineString and 6.1.11 Polygon types derived from them.
Whatever the final decision is, the next CDB standard should consistently use the same standard terminology for geometry in all places it is referenced in the standard. For interoperability, I suggest the GGDM terms since they are still consistent with Simple Features.