[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: Minutes of Pacific UBL TC call 19|20 September 2005
MINUTES OF PACIFIC UBL TC MEETING 00H30 - 02H30 UTC TUESDAY 20 SEPTEMBER 2005 ATTENDANCE Jon Bosak (chair) Stephen Green Ken Holman Kumar Sivaraman STANDING ITEMS Additions to the calendar: http://ibiblio.org/bosak/ubl/calendar.htm None. Liaison reports: Tax XML TC, ebBP TC Nothing happening this week. Subcommittee report: HISC GKH: No meeting this week (working on code lists); will be sending some stuff to ZR. Subcommittee report: SBSC SG: Nothing this week. Subcommittee report: PSC No meeting this week. Subcommittee report: TSC No one available to report this week. Team reports: Code List, Digital Signatures, Catalogue No one available to report this week. Review of Atlantic and Europe/Asia calls No comments. Schedule review: http://lists.oasis-open.org/archives/ubl/200508/msg00167.html JB: Our timeline for minor versioning has fallen apart due to scheduling problems among the NDR team. The focus this week is on code lists. ACTION ITEM REVIEW ACTION: SW to put out a f2f meeting notice to the ubl and ubl-psc lists. Done. ACTION: JB to ping PB regarding form choices and mappings. Pending. ACTION: TM to document the spreadsheet structure and format by the first week of October. This means a prose description of each column in the spreadsheets that we are now using. Pending. ACTION: TM and SG to examine the latest ATG2 draft and report back this meeting as to whether it meets our requirements. SG: Discussed this in email on the list. It looks like the ATG2 schemas haven't changed, so everything we've said in the past still applies. The only outstanding question is whether we need to create our own qualified DTs; my feeling is that we do need to have codes using amount to have supplementary component attributes, and this requirement is enough to justify adding UBL amount as a qualified DT. But I understand and could accept the view that we should keep to ATG2 amount. JB: Will put this on the Atlantic agenda. ACTION: JB to find the notice(s) regarding the SBS by Cover and Geyer and check to see whether the extended process model has been mistakenly characterized as normative. JB: Clarified this with SW. The notice in question is the current language I put on the TC web page; it says that we've adopted the extended process model as the basis for work in this area, but it doesn't say anything about the normative or nonnormative status of the process description in the final standard. CODE LISTS High points of the discussion: - There are no enumerations in UBL 1.0 outside of the files in the xsd/codelist folder. Nothing with a type of udt:codetype is an enumerated list; all the enumerated lists have types with (name) followed by codetype. - AH has provided a breakdown of the enumerated lists in 1.0: http://lists.oasis-open.org/archives/ubl/200508/msg00096.html - It appears that our UBL 2.0 Class 2 schemas (no enumerations) actually comprise two subclasses: those that are not enumerated because it is assumed that the values are provided by a code list maintainer such as ISO and those that are not enumerated because it is assumed that codes will be provided by the user. An example of the second variety is CountrySubentityCode, for which an ISO specification exists but is too large (upwards of 30 thousand entries) to be instantiated as a single code list. Another example is TaxTypeCode, which is different for every trading partner. We'll call these subclasses of Class 2 (that is, lists for which UBL does not provide the values) Subclass 2a (values provided by some agency outside of UBL) and 2b (values specified by users, or perhaps by no one). The mechanisms we come up with to deal with these two subclasses are probably identical, but it's important to be clear on the distinction. - A key question is how a user would add enumerations to a non-enum code list. A more complicated question is how trading partners would apply variations of a particular code list in different structural contexts. See, for example, http://lists.oasis-open.org/archives/ubl/200509/msg00091.html where under "udt:codetype" in the report for OrderCancellation are five unique non-enum parents in several different contexts. In the message http://lists.oasis-open.org/archives/ubl/200509/msg00086.html GKH suggests using schematron assertions to provide both kinds of functions. - In discussing this idea, it was observed that (a) it would require us to provide a lot of machinery, but (b) modifying code lists and developing different versions for different structural contexts (perhaps even in the same document type) is something that many users will have to do anyway, using ad hoc methods if none is provided by UBL. - In UBL 1.0, only simple udt:codetype codes have more than one parent (one structural context), but this is an artifact of the 1.0 design; it is possible to have more than one code element or attribute associated with a particular enumerated set of values. So we need to key on the combination of information item (i.e. the element or attribute) and data type. ACTION: GKH to write up the proposed methodology discussed in the call, including the distinction between subclasses 2a and 2b, and prototype the next version, keying on each datatype/information item pair. (He will be in a plane at this time next week but will provide all in mail.) - We need to discuss where we stand with the development of the genericode schema in this week's Atlantic TC call. Jon Bosak Chair, OASIS UBL TC
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]