Derek Dahlsad is a technical wizard and sharp designer. Self taught in most respects, he pulls a formal theatrical design education and part-time computer science courses into a skill-set that is neither purely artistic nor limited by technicality.

Back To Main

More Posts from Derek Dahlsad
December 2007
July 2006
April 2006
March 2006
February 2006
January 2006
December 2005
November 2005
October 2005
September 2005
August 2005

From Screen To Press, Via HTML and CSS
29 Nov 2005, 7:52:41 am

CSS is evolving to accept page-ready formatting, allowing for the web-designer or content-management program to format their documents for both the printer and the screen. From a convergance point of view, this isn't just a printer-trick: it is a way to change the way content-oriented businesses manage their info for a variety of media.

A List Apart has very rudimentary descriptions of next-generation CSS that allows for page formatting. The article, at first glance, isn't much more than demonstration of some new CSS tricks that impersonate Microsoft Word functions, but it does show a step in a very different direction for CSS and XML.

HTML, strictly speaking, is a form of SGML -- a metalanguage that doesn't have direct commands itself, but describes a formatting of the file's data. It's a bit of a high concept, but SGML is used as the basis for both HTML and XML -- the 'tags wrapped in math symbols' idea of how to mark data using the most basic, cross-platform encoding. Now -- HTML and XML aren't the only applications of SGML. Early (probably much too early, given the climate) in the world of markup languages, SGML was applied to creating electronic versions of text documents, for easy interchange. The Electronic Text Initiative is a big name, providing tutorials and explaining how real-world texts could be marked-up for online viewing rather easily.

Regardless if TEI's tagging method is used, CSS can be applied to any SGML file. CSS -- Cascading Style Sheets -- is designed to work with any SGML-based language, and can apply formatting descriptions to any document, when interpreted by an appropriate viewer. Your web browser is one of those -- technically, a web browser is nothing more than a SGML/CSS interpreter who only understands the HTML parts.

A List Apart's example takes one HTML document, apparently designed for the experiment, applies one CSS file to it, to transform it to a printable-format document via a PDF interpreter. Don't get too excited over this process: outputting HTML to PDF, nicely formatted, can be done just as fast in OpenOffice as it is to write a CSS file for it. What you should look at is the way CSS and SGML work together. SGML is merely data, marked with tags; CSS is the formatting. MSWord and PDFs integrate both into one file, but there's an advantage to having the two separate: formatting can be dependent on the application. When I go to change a saved book from hardcover format to a smaller paperback format, I have to go into my document and change all the page settings, scan through the document to make sure everything still looks OK, maybe tweak a few other settings. That's not to say that I wouldn't have to do the same for CSS -- but were I to have a starting point -- a 5"x8" CSS file, a 6"x9" CSS file, and an eBook CSS file -- the changeover would be much simpler because the first step would be to tie the SGML to a different CSS file, rather than modifying the SGML document directly. In fact, if I understood how the formatting is tied to the tags (invisible in WYSIWYG programs), I might not mess with indents and positioning in the same way.

This is all paperworld application, but, as I mentioned before, CSS and SGML are basic web-browser functions. Online publishing is no longer done with a HTML editor: content management systems are all the rage, from Blogger to Wikipedia, to take raw data and format it for online interaction. Online interaction is a separate process from data collection, which does not necessarily include content formatting. Often the data content is stored in an XML (a SGML subset) format for easy interpretation by software, or (at the very least) an easily interpretable proprietary format that could be converted to XML or SGML. This isn't an off-the-shelf change fixed with a few lines of code, but with the online text industry moving in line with the offline publishing industry, the Electronic Text Initiative (and their bretheren) might become a focal point for the next generation of HTML: if the bare tagged code is marked-up for both print and online viewing, producing a website or producing a book would be no more complex than printing your webpage at Kinko's. Online controls, like those used by Google Book Search, could prevent complete reading online -- but, with a push of a button, an online customer could buy a book containing the website contents, at that exact moment, by paying for the HTML to be matched up with the proper book-format CSS file and sent directly to the printer/order-fulfiller, regardless of the impermanency of the online information.

Let's think: what industries are obscenely ephemeral, rely on a print version for financing, but are struggling to make their online ventures viable yet harmless to their print aspect? Magazines and newspapers. Newspapers, long relying on internal content management, probably are doing this already, but in the reverse direction, from paper-print to online, using HTML and CSS for formatting, but there's a large crop of magazines that have grown out of the 'zine and webportal world. While the "Buy Today's Issue" button won't be useful until home printing-and-binding equipment is available, but making the web-to-print conversion is a cost-saver. Writing articles in a SGML format isn't new -- but what content management system uses it today? A subscription online magazine like could offer the tactile fun of a weekly magazine, taken directly from the SGMLized website, run through a CSS formatter for magazine layout, then electronically sent to a magazine printer & subscription-fulfiller. This eilminates the costly process of having a parallel-yet-separate 'hardcopy department' devoted to making the paper magazine. Of course, there would have to be skilled technicians managing and supervising the process, but the SGML/CSS connection would do much of the typesetting work.

On a very low level, insersecting SGML and CSS would produce a reason for lower-cost printing and binding. If BlogBinders recieved a SGML file, a CSS file, and the SGML DTD (for cross-program compatibility), they could instantly produce books from blogs, without any side-stepping 'shims' to make it work -- they could even provide the CSS files to fit their own specifications, and allow instant preview online. Zines would sprout up, thanks to open source and a crop of SGML/CSS3 based content management systems that can take a crudely-written zine and format it for distribution on the World Wide Press. Utopian as it may sound, HTML did this for the web: the rapid dissemination of content in a pretty human-readable format opened the internet to the general public. Print publishing is still a nation-tied industry with a long period of time from content creation to production of a hard copy. Using the web's quick-content-creation-and-formatting model, and applying it to the old-world computer-to-press process, books and magazines could be as easy to create as a blog is today.

Back To List