We went to Seattle hot on the tracks of XML tools. There we found that the biggest footprint is without doubt that left by Mozilla, and most of the attention rightfully goes to the future of XML in the browser. A collection of smaller prints, however, marks a simple, well-defined, and long-anticipated turning point in the evolution of tools for writing structured documents.

We report here on four products in development that left footprints at XML '98. The one that is closest to release is the alpha XED from the Language Technology Group. Of the other three, Interleaf's BladeRunner has a release "quarter" (no. 3, 1998), SoftQuad's XMetaL has a release year (1998), and Xerox's Raven is a proof of concept that does not yet have a development budget, although discussions are under way both within Xerox and between Xerox and some large strategic partners.

But the trend is not to be found in the individual tracks so much as in their overall pattern, and this is what we see: the emergence of a broader continuum of structured editing products.

A year ago, the continuum ran from high-end, SGML-structured editors to low-end, HTML, fixed-DTD editors. The scale was based on which type of structure and semantic markup was supported. But the overall category was called "structured editor" and that was sufficient to distinguish it from unstructured, wysiwyg word processing, from desktop publishing page makeup programs, and from database front ends tied to predefined data schemas. You defined your publishing requirements, then chose your tool: structured editor or otherwise.

Today, we have signs of further differentiation within the structured editing marketplace. The new continuum parallels that which has existed for some years within word processing and publishing: At the low end, we have editors facile at handling text but light in presentation handling, such as XED, which is designed for fast, hands-don't-leave-the-keyboard entry of well-formed text. You might think of it as the Notepad or Emacs for XML. At the upper end are the Frame and Interleaf products, structured editors that work in the context of complete page makeup.

Attempting to fill the hole in the middle are Xerox and SoftQuad, with new XML products on the drawing board, and Corel, with its attempt to revive WordPerfect. If Trellix were capable of outputting an XML data stream, it, too, would fit into this space. But the big player in the middle, by all rights, should be Microsoft, which is reluctant to travel far down the structured editing path.