RE: [opendocument-users] Alignment with MCE

["Dennis E. Hamilton" <dennis.hamilton@acm.org>]

Hi Rick,

Well, I am not going to pick on +1's from anyone [;<).

With regard to additions to namespaces, subsumption requires that there be a
new namespace in which all extended names in the subsumed namespace are
supported in exactly the same way as in the (version of the) specification
that introduces the subsumed namespace.

When subsumption is not the case, I find this statement important: "Future
versions of markup specifications SHALL specify new namespaces for any
markup that is enhanced or modified by the new version, which a markup
consumer of that version of the markup specification [wording problem here]
would include as an understood namespace. ... Regardless of whether a new
namespace subsumes a previously defined namespace, markup consumers based on
a new version of a markup specification SHALL support all supported
namespaces of the previous version unless the new version makes an explicit
statement to the contrary."  This is MCE section 9.1, paragraph 6.

I find this to be a marvelous provision for alerting a down-level processor
to actual presence of unknown up-level features and provision of advice
around how such features are to be ignored or (via mustUnderstand) required
that the down-level processor fail because the feature is essential.  The
level of accountability required, and honoring of interoperability in the
family, is remarkable.

 - Dennis 

-----Original Message-----
From: rjelliffe@allette.com.au [mailto:rjelliffe@allette.com.au] 

Sent: Sunday, March 08, 2009 20:57
To: 'ODF Users List'
Subject: RE: [opendocument-users] RE: office-comment text:id vs xml:id (ODF
1.2CD01)


[ ... ]

+0.5   I don't think MCE does forbid the known namespace being added to.
It is more about how to treat markup in unknown namespaces. In general, a
breaking change is better handled incremementally rather than a clean
break: think on-ramps and off-ramps. So the old element allows both old
and new values, with generating data of the old values deprecated
(obsolescent, and being removed in some future version) and a clear
mapping from the old to the new.  This corresponds to how applications
will actually work (i.e. accept any 1.n.)

In other words, the policy is not backwards compatability in the sense
that any 1.m<n document is acceptable as a 1.n document, but that only a
1.n-1 document needs to be acceptable as a 1.n document.

Cheers
Rick

---------------------------------------------------------------------
To unsubscribe, e-mail: opendocument-users-unsubscribe@lists.oasis-open.org
For additional commands, e-mail:
opendocument-users-help@lists.oasis-open.org