How Hot is this Metal?

July 14, 1999

Liora Alschuler

What an XML authoring tool or XML editor or XML word processor is supposed to do is still very much up for grabs. XMetaL moves us a small step closer to defining the expectations for a commercial XML word processor, but SoftQuad has not taken the large conceptual leap needed to define the application in the way in which WordStar did for word processing in the mid-1980s. In a field divided between Arbortext's Adept, FrameMaker+SGML, XMetaL and WordPerfect SGML, there is no hands-down winner. Each has unique strengths that make it the best choice in certain situations.

XMetaL's strengths are price, ease of setup and customization, a simple user interface, and support for a standard style sheet mechanism. There is every indication that SoftQuad will respond to what it learns about the market for an XML word processor.

SoftQuad has used its years of experience to make a good assessment of the market for XML word processing. Its belief in a division of labor in Web publishing that mimics print seems justified: Web designers want to create a framework and then give a tool to content creators that leaves that framework intact. To the extent that the Web becomes more database-centric, and certainly when authors are writing for multiple media, a tool such as this helps bring consistency to the markup for organizations large enough to set up such a division of labor.

To some degree, the perception of XMetaL's ease of setup and ease will ride on the wave of CSS acceptance as the common style language for XML. Although you can't do everything you might want to do with CSS alone, the level of support in the normal view is a good start for a first release. We fully expect to see more of CSS2 as the product matures.

The Holy Grail for structured writing tools has always been ease of use with document validation. You can't hide the rules and enforce them at the same time. SoftQuad has gone a bit too far, perhaps, in hiding the rules along with the complexity of the underlying syntax. It won't be easy to find the optimum balance, yet it strikes us that the market-winning approach is likely to be one that acknowledges that there are rules and shows how the rules apply to a nested hierarchy of document elements. On the positive side, SoftQuad has done a good job of separating the structure and content of the document from the rules about its appearance. While we need to keep that distinction, it would be good to integrate application of style more thoroughly into the interface.

It is disappointing that SoftQuad took a step back from letting folks work with well-formed, but not necessarily valid, documents. You can do this in HoTMetaL 5, which will let you write your own tag names. Arbortext calls this "free-form XML editing," and it will be in the next release of Epic and ADEPT. Although it hasn't announced release numbers or dates, Arbortext has demonstrated a simple well-formed word-processing feature that lets users quickly assign styles to new tags so that they can view or build a structured document without an associated doctype. The initial feedback on XMetaL included consistent support for well-formed writing, which is interesting coming in parallel with increased interest in support for the W3C Schema activity. With schemas, SoftQuad will be able to provide richer validation, including data types, which is closer to the business logic that people are trying to capture. Expect to see structure or tree views of the document back "by popular demand" and serious consideration given to editing of well-formed text.

Market response. We suspect early-adopters and those already using SGML editors are already looking at XMetaL. New markets will open up where significant numbers of writers need to share an agreed-upon document model. Since we expect that most of these users will be working with content management systems, it is critical that SoftQuad secure lasting relationships with a fair number of VARs and OEMs.

SoftQuad has taken a good stab at defining a new type of XML-authoring tool and we expect to see improvements as the market for specialized XML editing matures. Customers looking for a stand-alone, generalized XML word processor still have to prove to vendors that a market for such a tool exists.

Have you used XMetaL? What do you think of it? Share your opinion with us.