setRedraw002: 'add' or 'addHighlight'?

[Lofton Henderson <lofton@rockynet.com>]

Implementors --

At least one implementation does this ('add' vs. 'addHighlight') wrong.  Or 
at least an early-2009 version did it wrong.  Please check your implementation.

All --

It has been asserted that setRedraw002 might be wrong:

>[...]
>setRedraw002 ... I think the test is invalid. What was
>the agreement again about the keywords, add or addHighlight?

There was some discussion about this in January (mostly in the WG).  It is 
a slightly confusing situation, but (IMHO) the 2.0 standard is clearly 
specified (as is 2.1).  Here is my research and my conclusions.

2.0 SPEC:
-----
First, here is what the 2-1/2 year-old 2.0 spec says (and 2.1 spec as well):

>Ch.3 "IC" normative spec:  the keywords in fragment are newHighlight and 
>addHighlight.  If you generate a URI fragment (in a CGM-to-CGM link, in a 
>HTML-to-CGM link, etc) with highlight controls, you use those keywords.
>
>Ch.5 "DOM" normative spec:  if you want to control highlighting through a 
>DOM highlight() call, you use 'new' and 'add'.
>
>As long as you follow the 2.0 specification for highlighting in fragments, 
>and follow the 2.0 specification for highlighting via DOM highlight() 
>calls, everything is fine, as long as your viewer has correctly 
>implemented the specification.

setRedraw002 test (.htm):
-----
It uses 'add' in its highlight() method call, consistent with the 2.0 spec.

WG Resolution:
-----
[1] http://lists.w3.org/Archives/Public/public-webcgm-wg/2009Jan/0028.html

Since I find nothing more in the archive about this topic, I assume the 
resolution is "no change to 2.0 spec".  (Else, we would have had to 
generate a 2.0 erratum, yes?)

Conclusion:
-----
If one uses the DOM highlight() method, the parameter value to use is 
'add'.  On the other hand, if one embeds object behavior keywords into a 
URI fragment (i.e., #webcgm-fragment-stuff), the keyword is 'addHighlight'.

Thought or comments?

Regards,
-Lofton.