I expect that the following example would best be modeled as a supporting type and would like to get some feedback. I would like to use Freebase to summarize and store somewhat complex conditional values. For example, I would like to create a topic that holds the following information: In publication 1 there were 5 eggs of species 1 in habitat type 1 but in that same publication there were 4 eggs for species 2 in habitat type 2.
Currently I can model these 2 topics (i.e. pub 1, species 1, eggs 5, habtype 1 vs. pub 1, species 2, eggs 4, habtype 2) through a standard type with little problem. Data entry for relatively redundant topics will be relatively slow going until cloning is available but it is possible. It seems to me that a supporting type should probably be used to maintain these entries. Correct? I sure as heck don't want them coming up in the UI because they are too complex among other reasons.
This brings up the following question: Why should I give it these kinds of topics a name? If the names aren't used in the UI then it is just a system identifier. A time stamp would work just as well. Also, I probably won't know a good name for these topics until I enter their attributes. Maybe a name could automatically be created and then updated as a concatenation of topic properties? Either way, naming the topics for this kind of supporting type will often be pointless and a waste of time. Is it possible for a new feature to allow the checking of a box for automatic generation of names for a supporting type?
