Details
Sub-task
Closed
Blocker
Description
This is a summary of existing and proposed mappings from the basic types defined in ISO 19103 to the Java environment and the GeoAPI interface library. This document should be subjected to peer review and OGC vote.
Formalizes the proposals for Records and Namespaces.
Attachments
Issue Links
Activity
200Mb? The Pdf is only 0.5Mb. 
Besides, it'll do you good if you're planning on following coverage development.
I don't recall a dynamic typing system in 19115 (Metadata). Our feature model was designed without reference to the 19100, so one would expect some duplication. However, you've offered to spend some time integrating the new FM with Record. I cannot speak to "Catalog" specification, as I remain ignorant of it.
A caution, however, RecordTypes do not possess inheritance. They may be grouped into packages, but not into parent-child relationships.
However, w.r.t. code collapse, look at the big picture: OGC is steadily migrating their standards to ISO. This is a past (2005 was a big year), present, and future activity. As standards are migrate to ISO, they tend to become more vertically integrated. If you had 19123, I could tell you to compare the OGC coverage spec with the ISO one. But you don't, so you gotta wait till OGC publishes it. The upshot is that ISO, in general, does a better job of building on previous standards. So my answer would be that I foresee code collapse with every ISO migration.
Your interpretation of the <<Type>> stereotype is correct. I will have to revisit this issue later, however. Since you have OOo now, feel free to edit the document with change tracking on and email to me.
Besides, it'll do you good if you're planning on following coverage development.
I don't recall a dynamic typing system in 19115 (Metadata). Our feature model was designed without reference to the 19100, so one would expect some duplication. However, you've offered to spend some time integrating the new FM with Record. I cannot speak to "Catalog" specification, as I remain ignorant of it.
A caution, however, RecordTypes do not possess inheritance. They may be grouped into packages, but not into parent-child relationships.
However, w.r.t. code collapse, look at the big picture: OGC is steadily migrating their standards to ISO. This is a past (2005 was a big year), present, and future activity. As standards are migrate to ISO, they tend to become more vertically integrated. If you had 19123, I could tell you to compare the OGC coverage spec with the ISO one. But you don't, so you gotta wait till OGC publishes it. The upshot is that ISO, in general, does a better job of building on previous standards. So my answer would be that I foresee code collapse with every ISO migration.
Your interpretation of the <<Type>> stereotype is correct. I will have to revisit this issue later, however. Since you have OOo now, feel free to edit the document with change tracking on and email to me.
New revision of basic tools proposal. Completed Records and RecordTypes section. Reversed position on the meaning of "depth". The units and measure section is still incomplete.