I am a bit confused about the XML DOM. The DOM is an API allowing programs and scripts to dynamically access and update the structure and style of documents. So far so good. This means that once your documents is recieved for example by a browser and have been parsed, you have access to this documents tree and can manipulate it. But does this XML document and its parsed 'tree' just disapear once your browser goes on to the next document? I understand that this is a matter of how the vendors implement this, but I am trying to understand whether the recommandation from the W3C says anything about this, which according to me does not. An idea would be to build a 'big' tree of all the documents parsed, where for instance each document would be a subtree. In this way these document trees are being cached and can be accessed later. There is however the problem with dynamically generated documents, like search results etc, but that would be dealt with the same way that it's done in today's HTML and DHTML, by not leting it be cached (merged to the 'big' tree). comments?