Data Model Design Review		5/15/01 - 5/22/01
		========================		=================

Minutes:

* We need to ensure that it is clear to tool developers, what they need to
  do to integrate with this.

* What if the PI wants to use a single attachment for all of the URLs? Will
  we support this, and if so how?

* Why are RA and Dec not unified between phase 1 and phase 2?
  How can we ensure that PIs go back and correct/improve their target
  position during phase2?

* We don't have any visit collections. Visit collections should probably be
  separate from exposure collections.

* How will consistency checks be done?

* Should we annotate on the diagrams which properties are user entered
  and which are computed?

* We will document the philosophy behind the package decomposition into
  model/view and controller in order to ensure that additions are put into
  the correct package.

* We will check the naming of default implementations of interfaces. Default
  normally appears first in the name. We will also check the philosophy
  behind "Default" and "Adapter" naming.

* We should capitalize the first letter of acronyms, rather than spelling 
  them out in upper case - javaism.

* ToolDataMap should be of type Map, but the implementation can be HashMap.

* How is undo implemented? The swing Undo management may not be sufficient
  for our needs. We will look into this.

* Flux means spectrum, and this is generic enough to be in the APT package.

* We may need a Global Unique ID for TinaDocumentElements.

* InstrumentUsage needs multiple SpectralElements.

* You do not need to have a Phase 1 proposal in order to put together a
  phase 2 proposal. APT must be able to construct a phase 2 directly, as
  well as from a phase 1.

* Call the phase 1 RA and Dec, provisionalRa and provisionalDec to help
  indicate that they are not the same.
  
* We have APT generic classes knowing about HST specific classes in the model. 
  See ExposureExpansion, ExposureCollection. We will correct this.

* Add spectrum to the phase 1 data.