RE: [ubl-lcsc] Re: Position Paper on List Containers

[Chin Chee-Kai <cheekai@softml.net>]

On Mon, 1 Sep 2003, A Gregory wrote:

>>While I agree with much of what you say about looking forward, I must
>>disagree that we can ignore totally ubiquitous tools such as Internet
>>Explorer. 

Don't use the 4 container rule's means "ignore totally ubiquitous tools
such as Internet Explorer"???   I find your argument stretching
the limits of imagination.  

Each message is already a container (with business meaning).  Due to
the probably sensible use of global element/type (ok, Gunther might
strongly disagree), each such element within is again (forced to be)
a container of sort (with a meanings attached). There's no argument
against "containers" per se, but the meaningless containers that 
the 4 container rules require.  They're also ill-defined as to give 
name clashes and, the way they are stated, are at most ambiguous in 
design.  There's no "disadvantage" or problems about those lists
as shown in the paper.  But they have not brought out any advantage 
as claimed without proof.

Show some proof of efficiency.  I can change my mind if the numbers
argue for themselves.  And, if you argue based on subjective 
assessment of elegance, there's not much room for discussion 
except you can win if you happen to yell louder by quoting names.

Sorry, but I shall stop responding to arguments based on "this 
container view is better than that", because there's not much value 
I can add to subjective views of elegance and prettiness.




>>Most people in my experience do use IE5.* or 6 to browse XML
>>because it provides a collapsible view that many text editors do not.
>>Also, the existence of collapsible tree-view and folder-based
>>hierarchies in XML editing tools and document management systems is very
>>common.

No disagreement there.  That's looking at XML in general,
not looking at UBL-specific implementation.  By that reasoning,
all XML documents should have containers otherwise they're
"ignoring totally ubiquitous tools such as Internet Explorer"????

There're benefits of folding/collapsing tree-views.  At least
when browsing UBL instances as they are now (with or without
containers), the feature is very handy.  But when you do that
for the sake of doing that (as in implementing those meaningless
containers for the sake of folding), and force the other side effects
to start creeping quietly into other areas, then a simple
argument of not "ignoring totally ubiquitous tools such as
Internet Explorer" is not sufficent a reason.




>>(Jon's arguments about readability stands, if only because the IE is an
>>example of a much more common UI behavior around XML.)

If you design a model for the sake of UI, how sensible is that 
is left to people at large to judge.

Like I said, to each his/her view, and I won't attempt to change
yours.  You probably feel strongly in your views about the issues
as I do.  We have both spoken.




Best Regards,
Chin Chee-Kai
SoftML
Tel: +65-6820-2979
Fax: +65-6743-7875
Email: cheekai@SoftML.Net
http://SoftML.Net/