The impetus for XED came from the Language Technology Group at the University of Edinburgh. The group is working to develop a way to store document databases of curricula and bibliographic material in a way that it is easier to update. Henry Thompson, who presented the product to the crowd in Seattle, writes: "We each spend a lot of time keeping our cvs and publications lists up-to-date, and then every year someone has to cut-and-paste from all of them to get the necessary data for our annual report. We are looking toward a lightweight document-management approach to ameliorate this problem." An editor that helps put the material into well-formed XML is the first step in this direction.

The XED editor ferociously protects the structure that makes a document "well formed" (balanced start and end tags), but it lets users rename tags or create new ones at any point in the document. While XED does not validate against a document type definition (DTD), it does track document structure and content models to provide context-sensitive cues for element insertion. If a DTD is available, it can be seen, but not edited, in the lower frame.

XED operates on the "virtual document stream," that is, the text stream that would compose a document if all pieces were concatenated in a linear stream. In XML terms, that includes nonparsed character data (CDATA) and entity references ("<" for "<").

An alpha version of the editor for Windows 95/NT and for Solaris 2.5 is available for evaluation purposes only from http://www.ltg.ed.ac.uk/~ht/xed.html. A Macintosh version is promised in the "medium term." The underlying technology includes the Language Technology Group's XML tools, which support Unicode encoding, although the user interface, built on Python and Tk, currently restricts the user to eight-bit encoding. The software is freely licensed with a restriction on further distribution or component reuse.