I have a few ramblings, mostly from the xslt/xml application point of view. - DTD loading lots of ideas float around this, probably gonna hit the parser hard - Well-Formed External Parsed Entity parser hell as well, but this stuff is used by the XML freaks quite a bit - unparsed entities - Support for different data models nsIContent should hold enough information to implement http://www.w3.org/TR/xml-infoset/ http://www.w3.org/TR/query-datamodel/ http://www.w3.org/TR/xpath20/ These build upon each other, are there other data models we might wanna need? - What's an HTML/SVG/XUL document? Can this be implemented by aggregation? That is, can we take an XML document and morph that into a HTML document? That would come in handy at least for XHTML served as text/xml.

Coupla questions: 1) are there any draft documents with written plans for Gecko 2.0? 2) I heard that one of the items planned was to fiddle with parser/content sink so the parser is running on a different thread than chrome (which would allow DTD fetching/validation to take place without freezing the UI, if we had a validating parser). Any truth to that? I'm a little confused by your last point--what are you asking for? It might actually be worthwhile to separate XHTML and (tag-soup) HTML in content; that would probably depend on better code sharing among the content sinks (long-planned).

Mass-reassigning bugs to dom_bugs@netscape.com

Gecko 2.0 has long shipped and nothing in comment 0 sounds like something we'd want to do. Let's close this bug.