ÿWPC2 ûÿ2ö£B Zå ?0­I Meeting Notes X3L8 12/5-7/94 Bruce BargmeyerBruce BargmeyerÿÊ   ÿHP LaserJet IIIHP3.PRSXxþ6X@X@ÐЫXxþ6X@X@<åþ 9Z ‹.Courier New RegularXûÿ2(03|x ÐÐMeeting Notes X3L8 Special Interest Group Meeting Held at Information Technology Industry (ITI) Council February 21©23, 1995 ÐÐ Attendees: Bruce BargmeyerÁÁÁhh#ÁEPA Larry FitzwaterÁÁÁhh#ÁEPA Dan GillmanÁ¸ ¸ ÁÁÁÁhh#ÁBureau of Census Alan GoldfineÁ¸ ¸ ÁÁÁÁhh#ÁNIST Paul HawesÁ¸ ¸ ÁÁÁÁhh#ÁBellcore Tammy KirkendallÁÁÁhh#ÁNIST Joe LeahyÁ` ` ÁÁ¸ ¸ ÁÁÁÁhh#ÁSAIC William Kenworthey, Jr.Áhh#ÁSelf Karen KirkbrideÁÁÁhh#ÁSelf Judith NewtonÁ¸ ¸ ÁÁÁÁhh#ÁNIST Phong NgoÁ` ` ÁÁ¸ ¸ ÁÁÁÁhh#ÁSAIC Frank OlkenÁ¸ ¸ ÁÁÁÁhh#ÁLBL Marlene PalmerÁ¸ ¸ ÁÁÁÁhh#ÁFEMTO Bert ParkerÁ¸ ¸ ÁÁÁÁhh#ÁMITRE Meeting Report: The meetings took place all day the 21st and 22nd and the afternoon of the 23rd. The following reports the issues raised and decisions made. 1. Review of parts of ISO/EEC 11179 1.1 ISO/IEC 11179 Part 1 oÂÂRework the example in 6©1 to better depict the perceived intent of the recommendations from the Stockholm SC14 WG4 meeting. In the text make a reference to a discussion in an informative annex that describes how data elements relate to modeling constructs such as "Attribute". The editor will fix the discussion in Annex B. oÂÂRemove the material about "data composite" and "generic data element" and move the issues to the X3L8 metamodel. These topics are not addressed or provided for in any of the other parts of 11179. oÂÂPaul Hawes (past editor of Part 1) and Dan Gillman (assuming editorship of Part 1) discussed Denis Hill's comments on Part 1. Some of the issues were raised in the group discussion. Dan Gillman will get back to Dennis Hill.Œ oÂÂSeveral editing changes were suggested to Dan Gillman. 1.2 ISO/IEC 11179 Part 4 The final draft was sent off to SC 14 to forward for publication as an IS. 1.3 ISO/IEC 11179 Part 5 The SIG recommends to the X3L8 Plenary the following comments for DIS Ballot: General: oÂÂChange format to comply with ISO guidelines and to make the document consistent with Parts 3 and 4. oÂÂCapitalize "Registration Authority" throughout the document Specific: oÂÂClause 3: Limit definitions to those substantively used in Part 5. E.g., delete "domain"; definitions between all parts of 11179 need to be reconciled. The reconciliation should be based on the definition matrix in Part 1. oÂÂClause 3.2: Change definition to be same as in Part 3. oÂÂClause 3.12: Change to a definition of "Object class name". oÂÂClause 4.13: Change to a definition of "Property name". oÂÂÁ€°° ÁÁÁClause 4.1: 1) Change "registration authority" to "registration authority identifier" Rationale: It is specifically an identifier that is being discussed here. It is not the full name of the Registration Authority; 2) Change order to place "registration authority identifier" before "registration identifier". ÂÂÁ€°° Á oÁÁClause 4.3: 1) Change "Registered Data Element Identifier" to "Universal Data Element Identifier" and add sentence: This is a universally unique identifier for a data element. Rationale: Need to distinguish this from the components; 2) Reorganize into a subclause for each component©©registration authority identifier, registration identifier, and version; 3) reorder to the same order as in the initial bullet list; 4) the term "registration identifier" is ambiguous in this context and revision should be considered. oÂÂClause 5, item 2: Switch order of "registration authority" and "registration identifier". oÂÂClause 5, item 3: Change "have been" to "be". oÂÂClause 6: Remove the word "domain" from 4th paragraph. oÂÂClause 6.1.1.1: Change heading "Object Class" to "Object Class Component"; change first sentence to "An object class name can be". oÂÂClause 6.1.1.2: Change heading "Property" to "Property Component"; change "information concepts" to "properties"; add "for the enterprise" after the underlined word "complete"; revise "properties" subclause to break out "representation" component as a separate subclause oÂÂClause 6.1.1.3: Change "Qualifier" to "Qualifier Component". oÂÂClause 6.1.2, item b): Example may need to be changed, based on comment for Clause 6.1.1.2. oÂÂClause 6.2: Include statement: "Within a naming convention, it is possible to establish syntactic and lexical rules." ÂÂÁ€°° Á oÁÁClause 6.2.2: Add to list: permissible character set and case sensitivity (i.e., state whether or not terms distinguish between upper, lower, and mixed case). ÂÂÁ€°° Á oÁÁAnnex B: Last diagram should be revised per clause 6.1.1.2; move "NAME:" to be in front of the suggested name. 1.4 ISO/IEC 11179 Part 6 Decisions: oÂÂWe need a substantive section on the procedures of registration including the criteria for each registration state. oÂÂWe will not use Annex H hierarchy. Instead describe how RA's can work together by creating new RA which can assign a new UDEI. oÂÂWe should create an informative annex or tutorial that shows an example of the whole process, including intermediate steps. oÂÂSeveral suggestions for editing changes were offered. ÙÙŒIssues: oÂÂWe need to determine a scheme for assignment of unique identifiers. Will ISO 6523 work? Suggestions included "distributed" assignment of unique identifier, such as recording the exact time of registration creation or using Global Positioning Satellite (GPS) coordinates. However, such distributed schemes lack the assurance of having a list of registries. oÂÂDo we want to cover registration of Data element concepts? 2. X3L8 © CDIF MOU Alan Goldfine lead a review of the MOU and its status. A couple modifications were suggested. The SIG unanimously resolved that the MOU be presented at the X3L8 plenary session with a recommendation to approve it for signature by the Chair of X3L8. 3. Metamodel discussion Paul Hawes and Joe Leahy worked together between last meeting and this to develop two levels of the metamodel, a "conceptual" level and a "logical" level. The results of this work were distributed and discussed. oÂÂThe work between Joe Leahy and Paul Hawes led to some revisions of the metamodel. The revisions are up on the FTP server. In the work, several new issues were raised, they are now on the list of issues. Issues discussed at the meeting are listed below. oÂÂAgreement: We will continue on a concurrent course, resolving issues on the "conceptual metamodel" version (recorded by Paul), translating the "conceptual metamodel" version to a "logical metamodel" in IEF (Joe Leahy), and working the translation to CDIF (Tammy Kirkendall, working with Joe Leahy). oÂÂNew Issue: how do we deal with dimensionality (e.g., "dimension" is length) of property classes, and how do we enforce consistency between dimensionality of the property and units. Raised by Frank Olken. oÂÂNew Issue: review names for each part of conceptual and logical metamodels. E.g., "Label" in current conceptual metamodel could be called "Name". "Data concept" could be called "Data Element Concept". Raised by Judy Newton oÂÂNew Issue: Is there a difference (that we want to record) between a single concept data composite and a multi concept data composite? Raised by Dan Gillman oÂÂNew Issue: Does a Data Concept (Data Element Concept) require a distinct name(s)? Raised by Paul Hawes. Answer: Yes, the fusion of Object Class name and Property name would produce only a structured "classification name". There are many other possible names at this level. oÂÂNew Issue: The current metamodel does not allow compound datatypes, it probably should. Raised by the group oÂÂNew Issue: need to be able to use multiple taxonomies for each of the components (that have a taxonomy). How to represent this? Raised by Frank Olken oÂÂNew Issue: Does a Logical Metadata Class get a new Version Label when it changes Registration Status? Raised by Paul and Joe. Resolution: No. Registration status has no impact on the data values or data meaning. oÂÂDo we need to support "generic data elements" i.e., a paring of property and representation without object class? Raised by Judy Newton. oÂÂDecisions: How to forward the work on the metamodel.  ©Â` ` ` ÂANSI level send out press release as soon as SD3 is approved. Press release describes the effort and how to get a copy. Have available the front matter and the list of issues available as well as the two levels of metamodels. Target date: End of April ÁÁ ©Â` ` ` ÂAs soon as we have resolved the issues, vote out of X3L8 for ANSI public review. ÁÁ ©Â` ` ` ÂCurrent thinking is that the "front matter explaining the metamodel", the "Conceptual Model" and the "Logical Model" will all be part of the ANSI standard. ÁÁ ©Â` ` ` ÂIt is believed that the metamodel in its current state (particularly the logical level metamodel) is ready for translation into CDIF notation and integration into CDIF. Tammy will perform the translation and integration, referring to Joe Leahy and Paul Hawes as issues arise. As the work proceeds it will be brought for confirmation to the X3L8 SIG at regular SIG meeting. Email and other forms of communication can be used between meetings. Time frame: initial translation and integration by the end of March.Œ™ÁÁ ©Â` ` ` ÂIt is believed that the X3L8 metamodel in CDIF notation and integrated into the CDIF Integrated Meta©model, as derived in the above item, is sufficient to identify issues that need to be addressed the CDIF Integrated Meta©model. A list of issues should be available to L8 by the end of March. The SIG will consider these, and forward them to CDIF as appropriate. The following are keyed to the list of metamodel issues, which is kept as a separate document on the X3L8 FTP server. oÂÂIssue 3: Decision: We have created an associative entity, currently labeled "data concept", which inherits all the attribution of "logical metadata class" and has the following unique attributes? "Property Class Qualifier Name" and "Definition Text". This item is resolved. oÂÂIssue 6: Decision: This can be taken care of by the four attributes of "Registration Authority": Unique Identification Label, Natural Language Label, Address Description, Derivation Expression Language Label. Status: Resolved. oÂÂNew Issue: Does the qualifier of "Representation Class" change the meaning of the data element? Decision: No. Qualifiers that change the meaning occur at the Property level. At the representation level, the qualifier describes the representational form. E.g., Alpha or Numeric as a qualifier of Representation Class, "Code". Status: Resolved. 4. Next meetings: SIG Meetings©©the following SIG meetings will be held at NIST: Wednesday, March 29 Admin. Bldg., Room B111, 9:00 a.m. to 4:30 p.m. No dining room for lunch Thursday, March 30 Building 221, Room A366, 9:00 a.m. to 4:30 p.m. Admin. Bldg., Dining Room A Friday, March 31 Building 225, Room B157, 9:00 a.m. to 4:30 p.m. Admin. Bldg., Dining Room A Tuesday, April 18 Building 223, Room B307, 9:00 a.m. to 4:30 p.m. Admin. Bldg., Dining Room A Wednesday, April 19 Building 223, Room B307, 9:00 a.m. to 4:30 p.m. Admin. Bldg., Dining Room CŒThursday, April 20 Admin. Bldg., Room B113, 9:00 a.m. to 4:30 p.m. Admin. Bldg., Dining Room A Tuesday, May 23 Admin. Bldg., Room B113, 9:00 a.m. to 4:30 p.m Admin. Bldg., Dining Room C Wednesday, May 24 Building 225, Room B157, 9:00 a.m. to 4:30 p.m. Admin. Bldg., Dining Room C Thursday, May 25 Building 225, Room B157, 9:00 a.m. to 4:30 p.m. Admin. Bldg., Dining Room A X3L8 documents, including this meeting report, are available via Anonymous FTP from speckle.ncsl.nist.gov. Change directory to X3L8 and check the "readme" file for details.