Difference between revisions of "Context Data Model 1.1 Open Issues"
Revision as of 21:15, 27 April 2008
- BlankEntities and Entities.
- The current HOWL and the current IdAS don't make a distinction between an Entity and a BlankEntity. In the proposed HOWL update the BlankEntity term is introduced. Currently both it and Entity inherit separately from OWL:Thing.
- (a) Define Entity to be a subclass of BlankEntity: i.e. an Entity is a BlankEntity with a single literal EntityId attribute
- (b) Don't introduce this distinction and call them both Entiteis--most have a (single) literal EntityID attribute, some don't. --Paul.socialphysics.org 18:14, 4 March 2008 (EST)I'm leaning towards this one
- Can we represent closed (non-mixed) types in OWL so that the LDAP CP can represent its schema?
- Ability to declare user-defined Classes to be 'closed', that is instances of them should follow the class definition, but not include any other "extra" properties. (same as previous?)
- Closed or open simple data types
- Paul asserts that we have the ability already to specify a format constraint along with a data type. For example, one could say the data type of an attribute is normalizedString, but constrained to a pattern that looks like a telephone number
- We can do this by creating a Data Range. A Data Range has a base XML Schema type (e.g. string) as well as all of the XML Schema facets (e.g. pattern, etc.)
- Worth noting somewhere: Entity Relations are only an abstract super-types for real, useful relations like "memberOf", "reportsTo", "friend". Or "hasFavoriteBook", "hasCreditCard", etc.
- Tony: We don't have a simplified description of the data model
- Need a simple-to-follow set of pictures that explain the data model
- This PPT was updated to the latest concepts terms and improved a bit based on feedback from the Jan/Provo F2F: Higgins Data Model Intro.PPT
- This is a runtime data model, there are not yet any tools that can create the graphs that I think folks might need. Restating: The data mode is an un-typed mode, (no sub-classes) makes you look at each instance to determine its type, this is not suitable for data mining and creating graphs of the data characteristic. [from Tony 2/21 higgins-dev email]
- When using relations there is now way to tell what relation we are really talking about. [from Tony 2/21/08 higgins-dev email]. Restated by Paul in 2/21/08 email: Given Entity (E1) that has two Entity Relations emanating from it, e.g one pointing to E2 and another pointing to E3, then are you saying that we’re lacking a way to “tag” or otherwise distinguish between these two Entity Relations?
- Another "core" subclass to consider is Account (or something that represents the attributes needed to authenticate to a system). It is important to figure this out early in the subclassing, because many repositories represent accounts as additional attributes of a Person and other repositories represent the account as separate from the Person. In addition not all accounts are Persons, the account could represent a device, application, or a group. [David K-M email of 2/22/08 on higgins-dev]
- Mixed attribute value data types
- Daniel points out that it would still be good to pass type on each value add:
- Resolution: we can mix types.
- Can an attribute have mixed values consisting of both simple and complex?
- Resolution: Yes.
- Many same-types attributes
- Resolution: No. (We should document that this isn't allowed/possible)
- Allow zero-valued attributes
- Resolution: we should not allow zero-valued attributes in the model per se. It is true that for access control reasons, no value will be returned in some cases.
- LDAP Issues and To-Dos --open issues specifically related to LDAP schema