Re: [codelist] A first draft of a new informative annex regarding life-cycle metadata

["G. Ken Holman" <gkholman@CraneSoftwrights.com>]

At 2022-01-26 01:55 +0100, Andrea Caccia wrote:
> Hi Ken,
> Thank you very much and sorry to be back only now.
> The text looks fine to me and reflects our 
> discussion. Just spotted a typo in the last 
> sentence (â??divine" should be â??defineâ?? I guess).

We could change it, but I did mean "divine" as in 
the verb to discover something that is obscure 
... though I see now with a web search that the 
use of the word as a verb itself is pretty obscure.

So, yes, we can change it to "define".

> About the examples, I understand your point. If 
> we put a complete implementation of a code list 
> as example can create confusion.
>
> But what about adding just fragments of code 
> lists to show each specific concept?

I worry that such will be perceived as "the standard way to do it".

Let's talk about it on Friday.

. . . . . . Ken


> Andrea
>
>
> > Il giorno 9 gen 2022, alle ore 03:05, G. Ken 
> Holman <gkholman@CraneSoftwrights.com> ha scritto:
> >
> > Fellow CLRTC members,
> >
> > Please find below a first draft of text for a 
> new informative annex to consider adding to the 
> genericode specification. This annex is meant 
> to address the comment received during the 
> review process asking about universal concepts of life-cycle metadata.
> >
> > It is just a first cut, but I wanted to get 
> feedback in case any member thinks I'm missing 
> the point or disagrees with what I interpreted 
> from our last meeting. I was tasked with 
> drafting the ideas we discussed, then we can 
> polish this up into finalized text.
> >
> > Please comment on the appropriateness of 
> adding such an annex, and, if appropriate, your 
> suggestions of the kind of edits that would improve what has been drafted.
> >
> > As indicated moments ago on the list, I'm 
> considering adding an example, but I'm 
> reluctant to do so. I wouldn't want the reader 
> to think the example dictates how life-cycle 
> metadata SHALL be represented if they miss the 
> fact that it is but an informative annex with 
> an example. It entirely defeats the purpose of 
> trying to convey there is no standard 
> representation of life-cycle metadata if it 
> gets interpreted as more than just an example.
> >
> > I like that the metadata (column headers) of 
> Pim's referenced examples are in Dutch rather 
> than English. If we decide to include the 
> example (and I'm still leaning not to, but will 
> defer to the group), at least it would give 
> pause to the reader who might think we've 
> standardized something across all coded value 
> domains by using a language other than what is 
> used in the specification itself.
> >
> > I look forward to your feedback!
> >
> > . . . . . Ken
> >
> > Considerations of life cycle metadata
> > Informative Annex
> >
> > Data life cycle management relies on having 
> useful life cycle metadata with which to record 
> properties of data and histories of changes in 
> that data. Such life-cycle concepts readily 
> apply to code list values. A few example 
> concepts include but are not limited to dates 
> and times indicating when a value in a coded 
> domain is accepted to be used by users or is 
> considered a valid value, a signal of 
> deprecation indicating when a value may still 
> be valid but is recommended not to be used, and 
> jurisdictional constraints within which a value 
> may be used to the exclusion of other jurisdictional boundaries.
> >
> > Custodians of code lists should consider all 
> such metadata issues from the beginning of 
> planning and creating a new list of values in a coded domain.
> >
> > The flexibility of genericode to declare 
> columns of value-level metadata supports 
> describing life-cycle metadata in addition to 
> any other kind of metadata or other information 
> that may be associated with values in a coded domain.
> >
> > Per the semantics of genericode, all of the 
> information associated with an individual coded 
> value or part thereof is captured in the <Row> 
> element. Human-oriented information about the 
> coded value or part in general that is not 
> intended to be processed by a computer is put 
> into the <Annotation> child of <Row>. 
> Computer-processable information is put into a 
> set of <Value> elements as metadata. 
> Human-oriented information about a particular 
> metadata value is put into the <Annotation> 
> child of <Value>. Computer-processable 
> information is put into either a <SimpleValue> 
> child of <Value> for a monolithic string value, 
> or into a <ComplexValue> child of <Value> for a 
> value that, itself, is allowed to be described richly using markup.
> >
> > The semantics and constraints of each 
> metadata <Value> are described to an arbitrary 
> extent in the <ColumnKey> element identified 
> with the <Value> element's columnRef= 
> attribute. Custodians of code lists should 
> consider how richly described each of the 
> metadata columns serves the reader or user of 
> the code list when they inspect the content of 
> <ColumnKey>. It is entirely up to the list's 
> custodian to establish and to document the 
> semantics and constraints of code-level metadata values in <ColumnKey>.
> >
> > Accordingly, it is not part of genericode to 
> standardize any semantics beyond the semantics 
> of the genericode specification itself of how 
> genericode markup is used to represent values 
> in a coded domain. There are no standardized 
> value-level metadata semantics, and so, there 
> are no standardized life-cycle metadata 
> semantics to be leveraged by genericode users.
> >
> > While some life-cycle metadata semantics are 
> considered somewhat universal, each custodian 
> may have a different perspective of such 
> semantics that impacts on their decisions 
> regarding how each semantic is to be expressed, 
> constrained, and documented. Therefore, the 
> OASIS genericode technical committee does not 
> attempt to divine any standardized code-level 
> metadata of any kind, including data life cycle management metadata.


--
Contact info, blog, articles, etc. http://www.CraneSoftwrights.com/o/ |
Check our site for free XML, XSLT, XSL-FO and UBL developer resources |
Streaming hands-on XSLT/XPath 2 training class @US$125 (5 hours free) |
Essays (UBL, XML, etc.) http://www.linkedin.com/today/author/gkholman |