Publications present particular difficulties to Web sites due to their number of pages, unique formatting and layout, and navigation/paging requirements. In addition, most publications are created and maintained in a format that is “foreign” to HTML, such as word processing or desktop publishing software. Things that are taken for granted in many publications such as a table of contents, tables, graphics, and footnotes can be very difficult to recreate in HTML pages. Likewise, a single publication may have many incarnations on its way to the Web—from a word processing document (the “original”) to a desktop published document (the “printer’s original”) to a series of individual HTML pages (the “Web original”) to a PDF file (on the Web and in print). As a technology manager explained, “I actually happen to have somewhat of an example of that going on right now, this consolidated plan, this three-hundred page plan. They want to put it out—they had the version out there in PDF that was for public comment. Now they’ve gotten the approved plan ... And the question from the deputy for policy [was] ... what do I do [and] what do you want it in, what format? ... and I said, well, you need to get a PDF of it and we can put it out in PDF. If you wanted an HTML, you need to send us the Word document. The PDF can go out almost immediately once you’ve signed that this is ready to be posted. I said the HTML could take a week.”

XML/XSL provides perhaps its biggest benefits in its ability to better coordinate publications. Since all the content for a publication can be contained in one single-source XML document, the problem with various versions and formats of the “originals” can be alleviated (see Figure 6 on previous page). Likewise, the peculiar challenges posed by publications for a Web page, such as the table of contents and footnotes mentioned above, can be “programmed” into a single XSL file and then applied to all the publications encountered on the Web site.

Figure 6. Creating and Maintaining HTML Web Pages via XML/XSL Files

In addition, one of the biggest challenges in the publication process occurs within the workflow. In most publication processes, once the document leaves the content developer and is handed off for review and edit, control of the source document can be compromised. In addition, different actors within the process can perform various jobs, so consistency and integrity can be compromised. A program staff member from a large agency explained it this way: “There are bottlenecks [in the] process, whether it is a piece of paper or electronic, ... it’s got to go through all those hands. The nice thing [about using XML] would be that ... we would just give them their piece ... to review and they could say, fine, and move on. So in that way hopefully it would make things move a little quicker.”