These days, commercial publishing involves more than just ordering up a large print run and mailing it out to customers.
Companies offering anything from financial statements to discount coupons want to be able to send high-quality versions of the same document not only by regular mail, but also over the Internet, by email, or to small hand-held devices (PDAs).
That currently requires creating separate documents. Even if each document is created digitally, each has to be styled separately by a graphic artist using a separate set of tools. It’s a time-consuming and expensive business.
Now researchers at HP’s Bristol, UK, lab have created a technology to simplify this process. They call it Multi-Channel Publishing, and their goal is to make it possible to design and lay out a document once – but publish it in any format.
‘We envision the development of easy-to-use, fully automated tools and processes for customising single-sourced content for publication and delivery,’ explains Tony Wiley, Project Manager for the Multi-Channel Publishing (MCP) team.
In Multi-Channel Publishing, documents are broken down into constituent components – content, layout, styling and semantics – then optimised and re-assembled automatically, according to the requirements of each media type.
That’s just the first step. Different publishing formats require different-looking documents. A document that looks perfect on a PC with a wide, high-resolution display will appear differently on a PDA with a small, low-resolution screen, a printed page or email.
The way that pages break differs in each, for example, and the way the elements of a document – images, personal names and addresses, informational texts, graphs and sets of figures – are placed in relation to each other varies depending on the characteristics of the media chosen.
To accommodate this, the team has written software with a rule set programmed to determine how the information is kept together in each publishing format, or channel. As long as the basic elements of a document are laid out and ‘marked up’ in a master document, this rule set can determine what information goes where.
The Formatting Objects Authoring Tool, or FOA, would know, for example, to always put company letterheads and addresses at the top of a page, but it would allow the text in a letter to run onto a second page if required by the media being used.
Written by researcher Fabio Giannetti, FOA is a Java-based authoring tool that allows users to create document templates and styling information without having to write them in the XSLT or XSL-FO programming languages. (XSLT, or eXtensible Stylesheet Language Transformation, is used to convert XML to other formats, most commonly, to HTML for screen display. XSL-FO, or eXtensible Stylesheet Language Formatting Objects, is one component of the XSL language used to describe a format for XML documents.)
Giannetti sees FOA as a starting point for developing a commercial authoring tool for multi-channel publishing.
‘In the early days of the Web,’ he notes, ‘people were hand-writing web pages in HTML with word-processing programs like Microsoft’s Notepad. But then along came programs like Macromedia’s Dreamweaver that let relatively unskilled people create web pages without knowing HTML. FOA could be the basis of something similar.’
The team has made <a href=’http://foa.sourceforge.net’>FOA</a> available as an open source program at sourceforge.com where it has so far received some 16,000 downloads, making it one of HP’s most successful open source projects.
The Bristol team currently has the Multi-channel Publishing engine working in demo form.
‘We’ve shown it can be done,’ says Project Manager Tony Wiley. ‘Now we need to integrate the technology into the publishing workflow.’
There’s some work to do. The engine’s transformation of the basic document elements into the different finished pages, for example, is not yet fully automatic. For the PDA version, users still need to adjust the style a little.
Still, Wiley says, ‘the demo is light years ahead of where a lot of other people are.’
The best currently available software in the document re-purposing space, Wiley notes, does use common data to create repurposed documents, but it doesn’t work from a master document so it still requires that each page of each document created for each channel be crafted by hand.
If the researchers can perfect their software, the demand for it certainly seems to exist. One potential use: Creating highly customised communications such as financial reports, transaction statements and documents tailored for specific languages and cultures.
Multi-Channel publishing stores the basic elements of each document in databases, which works as well with variable data as with fixed. Thus one element in a document might be something that every recipient sees, such as a company’s address or a report on the performance of a mutual fund. Other pieces may be tailored to the recipient’s individual interests, such as financial account details or special offers for customers with specific spending or saving profiles.