[Application-profiles-ig] minutes of August 12

Thu Aug 27 23:32:58 BST 2020

All, I have completed the minutes of August 12: hackmd:https://hackmd.io/OH9muEUFSJGJo10W3V9wswgithub:https://github.com/dcmi/dcap/blob/master/meetings/2020/2020-08-12.dcap_zoom_call.mdIn it, I took the action to do a summarizing email, which this now is. I am going to refer to the Hackmd file line numbers (visible in edit mode) in this recap: RESOLUTIONS 95: RESOLVED: separate columns for constraint type and constraint 106: Constraint takes place on value for which we have defined a value type; and the value type is the type of the value in the instance data. ACTIONS 185 ACTION: kc to create a list where folks can mark datatypes (xsd datatypes) [DONE -https://doodle.com/poll/vz7snn4i8xexgfg4] DISCUSSION (bits - there's lots more) 110 Tom: Unclear to me whether we would need to have a value type like 'literal' 'uri' or whether we could mix that with notion of a constraint type. 115-154 discussion of RDF "value type" and specific data types; keep both columns for now, but use of RDF "value type" is optional for those who only care about specific value data types. Cannot be inferred because of bnode 157-185 Discussion of value data types and what should go into the column 175: language string issues 191-215: Discussion of constraints; agreement to make constraint and constraint type step 2 of our work My analysis of where we were on August 12: - We need a stand-alone value type column that is primarily based on XSD types - The RDF node types column is additional, and may be part of the extended model - Constraint and constraint type may not be things we can define, but we can show examples and they can be very flexible - For now, we postpone the work on constraints so that we can present the model kc -- Karen Coylekcoyle at kcoyle.nethttp://kcoyle.netskype: kcoylenet