Dr. Strangeweb

Or, How I Learned to Stop Worrying and Love the Blog

Back before Tim Berners-Lee developed the origins of the World Wide Web, Internet pioneers used ftp to get documents, Archie look up the various ftp servers, Gopher to browse for documents on remote servers in a folder like hierarchy, and Veronica and Jughead to index Gopher servers.

We also walked everywhere in the snow, uphill, and we liked it. I recall the first web page I ever hit: it was NASA’s, and I was using the Mosaic browser and a brand new 14.4k modem. It was a painfully slow experience, and there was little on the web worth the wait. I recall thinking, "this might catch on, but a lot of things are going to have to change first." They clearly did, because you’re soaking in it!

Berners-Lee took a fresh approach to connecting information together on the newly emerging commercial Internet; his new web quickly replaced those other Internet services, which are all but forgotten today.

The web offered an easy way to create documents that could reference any other document, anywhere in the world. Using the rapid development tools in NeXTSTEP, Berners-Lee was able to quickly develop a markup language, a transfer protocol and a content addressing system: HTML, HTTP and the URL.

People found ways to stretch the web far beyond anything Berners-Lee could have imagined. The web turned from a simple presentation of linked content into simple content with rich presentation. In retrospect, the design of the web had a couple flaws.

First off, URLs were written backwards. The top level domain should have come first; we shouldn't type www.apple.com/store, but rather com.apple.www/store. As it is, URLs are like a numbering system where the digits are in reverse order on either side of the decimal, rather than following an orderly path from most to least significant.

Ideally, we shouldn't be typing archaic domain names anyway. The simplicity of the system entrenched it into normal use, but we’re now plagued with phishing frauds, domain name squatters, registrar whores, and all the other schemers trying to peddle, park, and pilfer domain names as if they were common bicycles.

URLs have actually regressed in utility as static pages get replaced with dynamically generated ones, with URLs that expire seconds after they are generated, making them neither Universal nor Resource Locators. Attempts to create alternative ways to browse content, such as Apple's stab at the 3D HotSauce browser, didn’t go anywhere. We're stuck with command line addressing of the web like we're stuck with QUERTY keyboards.

HTML was the second flaw visible after a few year's hindsight. Armed with nothing more than the most basic markup tags, HTML was a poorly equipped to lay out pages of rich content. Its real strength was its flexibility in user interpretation: it can display content graphically, in a text browser, or via a screen reader for the blind.

However, there were severe limitations on absolute placement of graphics and other page elements within the original HTML specification. Mechanisms like frames, tables and, more recently, CSS, have tried to fix this, but are at the mercy of users' browsers to interpret them correctly and consistently.

As the web morphed from academic document linking into flashy, commercial driven applications, web designers struggled with various ways to display their content uniformly in every browser, bypassing the inherent flexibility HTML was designed to allow.

Many designers were also unhappy with the openness of HTML, which innately allows end users to change what they see, and repurpose the original designer's artwork, layout and other content as they see fit. As much as designers hate having their ideas "stolen," the open design of the web resulted in lightning fast development fueled by the widespread sharing of ideas.

The same people vilifying various DRM schemes for locking up content and stifling fair use are often ready to form a hanging posse flash mob if they discover their webpage layout has been sampled, duplicated and repurposed for another task. Open source acts as a two edge sword!

Macromedia (now Adobe) Flash is marketed to web designers as a way to create non-HTML content that is better suited toward presentation, and also allows for the restriction of end users to sample and adapt content, or modify its presentation. One side effect is that closed source Flash can encumber accessibility by the disabled or use on any alternative devices, like handheld browsers, which were not envisioned by the original designer. Boo.

In the end, the only one way to make HTML work for exacting presentation is to code it by hand. Tools like Dreamweaver or GoLive allow designers to shoehorn designs built for print media into HTML, but often do so by creating bizarre, fragile, and sloppy HTML. Other applications, such as Word, allow users to generate HTML versions of a standard document, with even less predictable and more byzantine results.

The problem really lies with HTML; it's a great way to lay out simple hyperlinked information, but a really bad presentation language for rich content. For regular people wanting to publish documents that look a certain way, this is a real bugaboo. It requires either getting knee deep in HTML, or relinquishing control of your layout to an automated content generator, which is inherently limiting.

Coming up next: a closer look at various options for building HTML, for people who don't care about digging around in HTML. In particular, I'll compare Apple's recently released iWeb with my experience in building and mangaging a web site in a text editor.