[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: Re: [ubl-dev] XML NDRs: UBL 2.0, 2.1, CEFACT 2.1, 3.0
> These fundamental designs go a long way to determining other > differences such as UBL using imports and not using includes > whereas includes are used in CEFACT schemas. > Perhaps I should clarify that. I think the reason for the UBL design was to allow custom use of substitution groups (which require that elements be defined at a global level and referenced from types rather than defined locally in the types themselves). This combines with UBL's modular design to contribute to UBL's use of 1) imports and 2) multiple namespaces and their prefixes. I would think that you wouldn't get to use substitution groups with a CEFACT set of schema modules. That said, the focus moved away from any recommendation of substitution groups deliberately in UBL 2.0 and the focus instead is on using the extension point facility in each document. I'm not sure whether there would be any way to preserve UBL 2 compatibility if ever the NDR changed away from this Garden of Eden design though, even if the use of substitution groups is effectively deprecated (in the loose sense of that term). With CEFACT NDR I think there was no such intention to allow for substitution groups in either customising or versioning the schema modules and the natural choice in this case seems to me to be the all global complexType definitions with local definitions of elements (other than document root element) - which I think is called 'the Venetian Blind' design. I did once try converting a set of schema modules from the Garden of Eden to the Venetian Blind design and IIFC it was OK that way but could not be converted back because of some information lost in the first conversion (best check on that in case I got it the wrong way around in my memory). ---- Stephen D Green On 19 April 2011 07:40, Stephen D Green <stephengreenubl@gmail.com> wrote: > Pim, > > A quick look at the CEFACT NDR (I've been out of the loop for > both NDRs for a while) tells me the main difference is still that > UBL uses the 'Garden of Eden' pattern of all global elements > and all global types > > e.g. for some fictitious XyzType > > <xsd:complexType name="XyzType"> > <xsd:sequence> > <xsd:element ref="ID"/> > <xsd:element ref="Name"/> > </xsd:sequence> > </xsd:complexType> > > whereas the CEFACT design is what I think is called > Venetian Blind and has types declared globally but elements > locally > e.g. > <xsd:complexType name="XyzType"> > <xsd:sequence> > <xsd:element name="ID" type="IDType"/> > <xsd:element name="Name" type="NameType"/> > </xsd:sequence> > </xsd:complexType> > > These differences carry on in version 3 of CEFACT NDR and > 2.1 of UBL NDR. > > These fundamental designs go a long way to determining other > differences such as UBL using imports and not using includes > whereas includes are used in CEFACT schemas. > > The codelist difference affects UBL 2.1 more than UBL 2.0 as > UBL 2.1 finally detaches the remaining CEFACT codelists that > were still tightly bound to the schemas in UBL 2.0. The CEFACT > codelists may still be used but are not tied to the schemas. > > Like I say though, I've been out of the development for some time > now and just follow the progress from a bit of a distance (day job > limitations, etc) but this is how it looks to me re the NDRs. > > ---- > Stephen D Green > > > > On 18 April 2011 16:40, Pim van der Eijk <lists@sonnenglanz.net> wrote: >> Hello, >> >> For a new project that I'm starting on I am doing some research on XML >> Naming and Design Rules. I am trying to understand how some of the recent >> specifications relate to each other and differ from past versions. I am >> hoping some people on this list can help here (and I'm assuming more >> subscribers to this list will be interested in this). >> >> The September 2010 draft UBL 2.1 has an Appendix G on the naming and design >> rules used for UBL 2.1: >> http://docs.oasis-open.org/ubl/prd1-UBL-2.1/UBL-2.1.html#NDR >> >> Previously, in December 2009 the UBL TC approved NDR 2.0 as a Committee >> Specification. >> http://docs.oasis-open.org/ubl/cs01-UBL-2.0-NDR/cs01-UBL-2.0-NDR.pdf >> >> 1) What are the changes (if any) between this CS and the UBL 2.1 Appendix G >> NDR ? >> >> In August 2007 a comparison was posted between the CEFACT version 2.0 NDR >> and the (then-current) UBL NDR: >> http://lists.oasis-open.org/archives/ubl/200708/msg00034.html >> >> In the mean time CEFACT has published a version 3.0 of its XML NDR >> specification: >> http://www.unece.org/cefact/xml/xml_index.htm >> >> 2) What are the changes (if any) between the CEFACT version 3.0 NDR and the >> latest UBL NDRs ? >> >> 3) What are the changes (if any) between >> this version 3.0 CEFACT NDR and the earlier CEFACT version 2.0 NDR ? >> >> Pim >> >> >
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]