Real World Computing
Content is king
The advantages of data-driven generation over static HTML publishing are fundamental, but so are the problems. While Dreamweaver, with its excellent support for server-based scripting languages like PHP, ASP, JSP and ColdFusion, can certainly be used to create data-driven sites, it's much harder than writing declarative HTML markups. Most designers can knock up the occasional server-side script, but writing and debugging thousands of lines of custom code to query and update a SQL database, manage user sessions and all the rest requires a professional developer. This chasm between designer and developer began to split the web into tiers: a lower tier of those near-ubiquitous, but increasingly inefficient sites built from static HTML pages, and an upper tier of comparatively rare, super-efficient, enterprise-level web apps that dynamically generated their pages on demand.
Falling off a blog
Around ten years ago, though, a surprising new force upset this two-tier web. It hardly looked revolutionary: the word "blog" referred to any kind of regularly updated online diary. These were originally created using traditional static web-authoring tools, by obsessives with a lot of time on their hands, but in late 1998/early 1999 everything changed with the arrival of the first server-hosted data-driven tools, such as Open Diary and LiveJournal. What made them so popular was that the interface to their underlying database was through a simple browser-hosted web form, so anyone could create and edit their own web pages. Blogging turned the browser back into the free, seamless, simple web-authoring app Berners-Lee had originally intended, with the enabling intelligence provided by the server rather than the client.
The torrent of blogs that was unleashed demonstrated the significance of such easy content creation, but - along with most designers - my initial response was "so what?" This was partly because the first text-only blogs looked dreadful - a failing quickly addressed by adding the ability to upload inline graphics and some sensible application of CSS. A more fundamental design issue was the way blog content was organised: the automatic organising of content into a chronological stream might be ideal for a diary or a train of thought (say musings about Britney), but it's wholly inappropriate for most general websites.
This major failing was addressed by the introduction of content tags. When posting this column, for example, I might apply tags like "web authoring", "RIA", "Dreamweaver", "browser" and so on, which could be used to automatically assemble any number of parent pages that list all items associated with each of the tags, or work out which other site items most resemble the current page and automatically generate relevant
While traditional web page designers could still achieve equivalent link-based accessibility for their static content with a lot of extra work, blogs popularised another kind of accessibility that's beyond Dreamweaver and its ilk: the RSS (Really Simple Syndication) feed. RSS is a dynamically generated reworking of data-driven content, delivered to the browser not as HTML but as XML. With RSS, end users can subscribe to your feed to automatically receive a list of all new content added to your site, delivered directly into their RSS reader. By forging an absolutely direct link between new content and interested readers, RSS-enabled blogs have opened up an entirely new way to get your message across, one that could soon become as significant a traffic source as the search engine. It's certainly a route that good designers simply can't afford to ignore, and to me represents the final nail in the coffin of traditional static HTML web publishing.
