Toby --
The Jira project is out of date - the most recent WD is WD28, and
they all show as unreleased. Please update when you get a chance.
Two notes to you in 3.5 and 3.6 below.
All--
Rather than post Jira items to an antique Working Draft here's what
I promised from today's meeting.
- UML diagram and proposed note was sent to the list around 90
minutes ago and is in the archive under Contributions. Public
link is
https://www.oasis-open.org/committees/download.php/54201/Obj%20and%20Subclasses%2020140925.png
for the post-meeting version.
- UML Enterprise Architect File (EAP) available on request to
any TC member that would like it. It would be confusing to post
it.
- In WD32 pdf I note the following:
- line 522 the link showing the URI is a live link; remove
Word hyperlinks in examples.
- Line 869 ditto
- Line 2436 No update for what changed in WD32
- obix-v1.1-wd32-changes.pdf does NOT show changes. It seems
the same as obix-v1.1-wd32.pdf, the clean version.
- We should be packaging the current schema/schemas with the
spec, I think - Toby? We'll need that for public review, can
live without it for now as long as the schema(s) have the WD
number.
- Again Toby - can we name the specification artifacts
obix-v1.1-spec-wd32? Don't know what the starter doc was
called.
- General: the internal/external links in WD32 are very well
done - I've worked with a lot of OASIS and other specs, and
Craig has done a particularly good and consistent job. Thanks!
- I cannot create a valid XML artifact based on the schema due
to the absence of "obix:Nil" and confusions around "obix:obj",
the latter as a shorthand for something of type "Obj". We
discussed this in the meeting.
- Address the "well-known objects" and the perhaps related
"lobby" issues. This is partially related to the item above. In
my opinion we want to think carefully about what we add to the
standard, and how extensibility is preserved, while maintaining
a standard base that can in fact be extended by implementors.
Perhaps "Lobby" should be an abstract class (UML)/ abstract type
(XML) that MUST be made concrete by implementors?
- The shorthand "obix:Nil: for "an object of type Obj that has
Obj.null set to 'true'" seems clumsy, but the existence of a Nil
object makes sense to me to complete the logical space. Needs
further discussion and thought. In effect, "Nil" is a literal in
the space of Obj's - so I'll need to look at examples of use to
have an opinion.
Thanks!
bill
|
