XML, XLL, and XSL: Current Status, Next

By Bob DuCharme, XML Correspondent

The biggest news at the recent SGML/ XML '97 conference in Washington DC was that the W3C, an international group created to develop common World Wide Web protocols, has formally issued the XML 1.0 specification as a Proposed Recommendation. After officially being in a Working Draft state for a year, it should only remain a Proposed Recommendation until the end of January, which gives the W3C members enough time to vote on whether to make XML 1.0 an official Recommendation, return it to Working Draft status, or to abandon it. To quote the W3C's home page, "The W3CXML Working Group has determined that the XML 1.0 specification is stable, contributes to Web interoperability, is supported for industry-wide adoption, and is ready to enter the review and voting process by all 229 W3C Member organizations."

In a session titled "Straight Answers on XML" on the first day of the conference, members of the W3CXML Working Group pointed out that while some have complained about their changes to the XML specification, there never was a specification until now: their changes were part of the work that made it an evolving, "working" draft. While W3C members can vote that XML 1.0 deserves Recommendation status only after additional changes are made, the specification looks pretty stable at this point.

The XML "Family" of Languages

W3C Working Group member Eve Maler explained that just as XML was designed to be a simpler version of SGML, the eXtensible Stylesheet Language ( XSL) is being designed to be a simpler version of the DSSSL style sheet language and the eXtensible Linking Language ( XLL) is a pared-down linking specification owing much to HyTime. While the group currently refers to the three X?L languages together as the XML "family" of languages, they seem interested in finding a better name to refer to the three specs as a whole. For current versions of the specifications, see http://www.w3.org/TR , a web page that includes many other W3C documents organized by their status as Recommendations, Proposed Recommendations, Working Drafts, or Notes.

XML Phase II: XLL

XLL, whose specification is titled "Extensible Markup Language ( XML): Part 2. Linking," is the second deliverable from the XML Working Group. Currently in W3C Working Draft status, the most recent draft of this specification is dated July 31, 1997. The Working Group had hoped to finish a new draft for the Washington conference, but work on part 1 of the XML draft moved this back to a new tentative target date of April, to be ready for the 7th International WWW Conference in Brisbane, Australia.

What's in the XLL spec? According to the current draft's abstract, "It is a goal to use the power of XML to create a structure that can describe the simple unidirectional hyperlinks of today's HTML as well as more sophisticated multi-ended, typed, self-describing links." More sophisticated links means providing for more categories of information that describe a link source and greater possibilities for describing a link's target.

More information in a link source lets a processing application determine from the type or purpose of the link how to implement it. It might be implemented as a pop-up, as an included file, in a secondary browser window, or the HTML ways: by replacing the link source's document with the link target's document or by jumping to a specific point in a document identified as a link destination by the appropriate element and attribute. XLL takes from HyTime the goal of describing different links by their purpose (for example, a footnote reference, a citation, or a glossary entry) and to then keep the implementation details of each link type separate from its meaning, much like SGML keeps the purpose of each element of text separate from its presentation details.

In addition to giving you greater flexibility in describing link sources, XLL offers many powerful new options (well, new to the HTML user--researchers involved in serious hypertext have known about these things for a while) in describing link targets. While an HTML A HREF link links either to an entire document or to a specific point in a document identified by an A element with a NAME attribute, XLL offers ways to link anywhere you want within a target document without requiring any point in the target document to be identified as a link target. It describes syntax that lets you say "link to the third bulleted item in the second bulleted list in the fourth chapter element," or even the third letter within that element. You can specify two specific points with a target document as the beginning and end of your link target.

XLL offers some very interesting new things to SGML users unfamiliar with HyTime. Although XML itself is a simplified version of SGML, and XSL offers a standardized way to do many things that SGML users are familiar with from DSSSL or various proprietary formatting systems, the added power of XLL-assuming that major browsers support this power-will offer much that is genuinely new to most document designers and their end users.

XML Phase III: XSL

XSL is the Working Group's Phase III deliverable. The current version of the XSL specification is titled "A Proposal for XSL," but as section 1.1 tells us, "It is not meant to represent a completed proposal for XSL." It was submitted to the W3C on August 27, 1997, and is now in "Note" status, which means that it's not even a working draft yet.

The proposal describes a stylesheet language that allows formatting of elements to be based, among other possible criteria, on an element's position, ancestors, and descendants. It also provides for the definition of reusable macros, generated text, the use of a scripting language for more sophisticated tasks, and many other powerful features.

Although it's being designed as a simpler version of the DSSSL, itself an ISO standard ( ISO/ IEC 10179:1996), XSL doesn't at all resemble DSSSL, whose roots in the Scheme and LISP programming languages mean that it groups related programming structures by using lots and lots of parentheses. XSL looks like XML, in accordance with its second design goal: " XSL should be expressed in XML syntax." The fifth design goal is that " XSL will be a subset of DSSSL with the proposed amendment" (that is, once the DSSSL specification is amended to make it a proper superset of XSL, just as proposed appendices K and L to the SGML 8879 standard will make it a proper superset of XML) and although XSL and DSSSL don't look similar, they're structurally similar enough that Henry Thompson has already written a program called xslj to convert a XSL stylesheet to a DSSSL specification (see http://www.ltg.ed.ac.uk/~ht/xslj.html ).

As people look for more sophisticated ways to format their web documents, they'll also notice the growing popularity of CSS, a web stylesheet language developed at the European Laboratory for Particle Physics that became a W3C Recommendation in December of 1996. XSL is not a competitor to CSS; design goal six states that "A mechanical mapping of a CSS stylesheet into an XSL stylesheet should be possible." In other words, soon there will be programs that convert any CSS style sheet to an XSL one.

Why convert a CSS style sheet to XSL? CSS stylesheets can do some quick, nice formatting with very little coding and can make an ideal starting point for formatting a particular class of documents for web delivery. However, as an online publishing system gets more complex, a developer trying to squeeze too much out of CSS is bound to eventually hit a wall. For example, CSS offers no way to change the order of elements, which is crucial to creating customized publications out of subsets of a large collection of SGML documents.

Although XSL has the furthest to go of the three parts of the XML specification, the fact that five of the proposal's eleven authors are from Microsoft and three are from Inso shows that important software companies are serious about making XSL implementation a reality. Browsers that support CSS formatting of XML documents may be ready first, but by the time people start clamoring for CSS extensions to make it more powerful (as they have for years with HTML) they won't need these extensions--they'll have XSL.

As with all useful but unfinished standards, the XML family of specs' authors are hearing exhortations to pick up the development pace, but I think that patience will pay off. If we give the W3CXML Working Group the time they need to do this properly, we'll all be amply rewarded. <end/>