July 09, 2010, 10:46 AM —
When they built the interstate highway system in the US, drivers immediately got easy access to almost very part of the country. Instead of weeks to travel coast to coast, journeys would take just days. Four lanes of shining concrete and asphalt could get the average family from New York to California without a single stoplight.
Of course, we also got traffic jams, dying small towns, and ubiquitous interchange villages out of the deal, too.
It's with a little trepidation, then, that I watch the rise of HTML5 on the Web. While the new markup language will mark significant progress for the Web overall, I can't help but think that some existing elements of web publishing aren't going to be adversely affected by the latest version of HTML.
To be sure, HTML5 promises a wealth of great new features. Web video's going to be easier to deploy, location-based apps will be even more robust, and plug-ins fade into the background--making browsing a more secure and efficient operation.
So... web site owners will now have geolocation data on anyone hitting the web with a GPS-enabled device (which will be nearly everybody); crackers will figure out new ways of exploiting web sites for all browser users, not just those with a vulnerable plug-in installed; every web page will now have some sort of video content, potentially creating a huge data bottleneck and inflicting eye-searing videos on unsuspecting users that will make <blink> tag abuse look like a deVinci masterpiece.
Am I being too pessimistic? Perhaps. But progress, especially in technology, has always come with a certain pricetag that typically catches us unawares. That does not mean we should stop progress altogether, mind you: but careful consideration of all the effects wouldn't be a bad idea, either.
My problem in this case is not with the HTML5 tools per se, it's what people will do with them. HTML, if you recall, was originally a content-driven markup language derived from the Standard Generalized Markup Language (SGML). In fact, HTML is just SGML with a different document type definition. At this level, it's helpful to remember that HTML was intended to be used as a document-centric delivery system.
Components of each document were to be well-defined, and thus easier to search, index, and design.
Now, though, with HTML5, the paradigm seems to have shifted. It's no longer just about the content-as-written-word. Now it's about content-as-anything. And "anything" is much bigger than you think.
Web 1.0 was about text and images generated from a central server. Web 2.0 was text, images, multimedia, and active data generated from a central server. Web 3.0 will be about all of this, but generated from anyone, anywhere.
That's the new paradigm that HTML5 and similar tools will give us. We're almost there now: microblogging, videos, images can be upload by anyone in seconds. HTML5 will just make it even easier. Presentation of content will give way to pure content generation, with less regard to how the content looks and (more worryingly) how accurate is the content.
If it sounds like I am kvetching about this, I will admit to some of that feeling. "Traditional" writers have a little bit of a hard time imagining a world where everyone and their brother, literally, will be able to publish content. My colleagues Scott Koegler and Preston Gralla have recently articulated this particular issue very well.
Even though my livelihood may be impacted a bit by a deluge of new content on the Web, I have a general feeling of optimism about a new interactive content paradigm with more of a hyperlocal focus. If anything, I want more user participation, if only to keep the Web from devolving into a commercially-driven entertainment channel for the gluttonous appetites of a sheep-like consumer base.
If HTML5 will make such participation easier, then bring it on, traffic jams and all.