Re: ActiveNodeSet/StaticNodeSet alternative

Curt Arnold wrote:

>Okay, maybe it will all come to me when I get some time this weekend.
>
>I don't see a use case for ActiveNodeSet that justifies the apparent
>complexity.
>
>For any moderately complex XPath expression, it would seem to be very
>difficult
>to detect whether a document change would invalidate the NodeSet without
>recomputing
>the results and comparing it element by element or just throwing it away
>if any modification occurs.
>
According to the spec, you do not try to detect whether or not the 
change invalidated the nodeset.  Any change to the document invalidates 
the node set, whether it needed to or not, so there is interoperability 
between all specs, rather than having some smarter or stupider than others.

Ray Whitmer
rayw@netscape.com

Received on Friday, 6 July 2001 13:43:23 UTC