This last sentence troubles many. Is it really necessary to introduce the behavior of the XML Processor in order to explain how default attributes work? Some have argued that the spec would be better if it simply defined carefully what the data in an XML document is; one of the things that affects this is whether or not attributes are being defaulted.
This argument has something to recommend it, particularly in terms of precision and conciseness. On the other hand, many find the explanation here, which in fact does describe what any reasonable program has to do about attributes, an easier explanation to grasp. This remains an open issue.
Back-link to specCopyright © 1998, Tim Bray. All rights reserved.