Summary of conference call June 22, 2004 In attendance: Larry Flon Hemant S Kolwalkar Brian Collins Andrea Westerinen Susan Malaika Tom Roney Concerning authorization and/or authentication: Larry argues that within the context of a database, if there is a provider that dumps table/index information into the CIMOM, then anyone can see the information. According to Andrea, there will be a standard mechanism in CIM v3 for all CIMOMs to enable granular access control. Version 3 is due out by the end of the year. How detailed do we want model? How much of the CWM model should we bring into this model, understanding that the more classes we add, the harder it's going to be to push it through the DMTF. And, the more complex the class model for the logical schema in CIM, the harder it's going to be to get the database providers to populate those classes. There was a lengthy discussion about whether or not we wanted to model user-defined types, and whether or not we wanted to model from the CWM work, and/or whether or not we wanted to support the SQL99 standard. After the discussion went a little further, into other issues in the minutes for GGF11 session 2, including stored procedures and triggers, the question is still at what level of specification do we want to model? We agreed that we need another pass at the use cases before we can push forward on the model, but then Susan joined the call and, after some questions concerning CIM (answered by Andrea), and after some further discussion about what we use as a base model, Susan mentioned that she had models (IBM) for much of what we were discussing. Susan will contribute UML models that IBM has for tables, columns, stored procedures, triggers, etc. It's either this or we use CWM. Susan will look into what they can contribute in the way of UML, and we agreed to work in a piecemeal fashion. Susan will get us the pieces as we need them. ACTION ITEMS: Susan and Hemant will talk about what they can send us to jumpstart this process (in the way of UML). Focus on the model bottom-up for now, and we'll see if we need use-case work later.