Re[2]: [cgmo-webcgm] Highlight test

[Lofton Henderson <lofton@rockynet.com>]

At 05:04 PM 6/14/2006 -0400, Benoit Bezaire wrote:
>I'm not so sure about this. After a quick look at our implementation,
>I don't believe highlight is implemented as an Style Property.

Okay.  But just to be clear (Don's later comment), there is no highlight SP.


>As far as I can see, we agreed that highlighting is a 2 state system;
>on or off (i.e., the API has two states: true or false). If we wanted
>a Style Property like behavior we need three states (true, false,
>inherit).
>
>So in that particular test file, the code is trying to remove a
>highlight which was never implicitly set.
>
>I also think that the DOM behavior and the Object behavior should
>match. I don't see anything with regards to inheritance of
>highlighting in the object behavior section:
>http://docs.oasis-open.org/webcgm/v2.0/WebCGM20-IC.html#webcgm_3_1_2_4
>
>I'm not convinced there's a need to have highlight behave like a Style
>Property... Those are my thoughts, feel free to comment.

Okay.  Let me ask and clarify a point then.  If you highlight() an APS, 
then its entire content is subject to the highlight, including an nested 
APSs, right?  So we don't really need to worry about inheritance model.

I can live with that.  I guess the description of highlight() makes it 
clear enough, that it's like a virtual #id(someObject, addHighlight).

Maybe that would be a good clarification comment -- that 
highlight(someObject) is identical to:  #id(someObject, newHighlight) or 
#id(someObject, addHighlight).  Which one?  I'd say the 2nd, 
'addHighlight'.  Thoughts?

-Lofton.