Cathode Tan - Games, Media and Geek Stuff
logo design by man bytes blog

Thursday, August 19, 2010


Yeah, people - this is what it takes to get me blogging again. Not one - but TWO stories that have mostly caps in the titles and begin with OMG. Because apparently this is such a slow news month that people really need to get the attention grabbing headlines.

Wired has declared the web is dead (though, to be fair, they've branched this out to multiple articles to offer a rather interesting and broad discussion).

First, never trust an article that starts with a graph - which I've now stolen and duped here:

It's an interesting graph ... but the important bit is to note that it spans 1990, when even the Net was barely a mote in someone's eye, to 2010 ... when high school kids have smartphones with more sophisticated browsers than most corporate networks. Since the graph is depicting total net traffic, and it is safe to assume that total net traffic has increased dramatically in two decades ... lines sliding down may actually indicate portions which have stayed even while the larger pie gets bigger. Note for instance that video doesn't even appear on the scene until the late 90's, and it wasn't until about 2005 when technology and bandwidth allowed creations like YouTube to flourish that it even begins explode.

So yeah, it's an interesting graph. But in relation to the argument at hand ... it's kinda bullshit.

The argument in short, and it's hard to put this in short terms with it several pages and oddly being split into two distinct articles running side by side (which is probably the kind of design decision which could kill the web) is that applications are sprouting everywhere to digest specific points of data as opposed to an army of browsers trolling for everything on the planet.

And I'm not really about to deny that argument, for in doing so I'd have to pretty much ignore the fact that the iPhone and the iPad exist at all.

For the sake of brevity, I'm going to encapsulate the argument around this table (also blatantly stolen):

Because it seems pretty core to their concept and is wildly misleading.

Browsers Versus Apps
This is a cyclical argument as old as DARPA and is slowly spiraling into oblivion. Firstly, it (and much of the article) ignores the fact that apps have made a resurgence because of concepts like REST which allows the web to be a generic source of data. Is it honest to say that a web server which is returning JSONP or XML based on a simple set of incoming variables is ... no longer a web server? The trend, even since the referenced (and failed) push revolution of the late 90's, has been for the formatting of data which allows nearly any client - native or web - to consume data. That Facebook is both a very successful website and iPhone app is indicative of this, and Facebook's recent Graph API allows for others to create clients of nearly any variety as well.

Also, I'm always surprised by the number of articles proposing this trend as fact while overlooking the fact that many major web publishers are foregoing native iPad apps in lieu of HTML5 ones. The real transformation of the web is into more than just a content center, but a data center as well, data which can be accessed by a wide variety of clients.

Syndication Versus Subscription
I'm actually not even sure why this is a line on the table. Syndication is the act of formatting to data so that it can be subscribed. This isn't an either or - it's a cause/effect. Moving on.

Update: It occurs to me the models at conflict here are subscribe versus follow, which are actually different metaphors. Insert question mark.

Free Versus Freemium
While catchy, this is also an apples to advertising based models comparison. Take Pandora for example, which embraces freemium but also pushes ads for the free version. As the name assumes, freemium is an extension of a free (usually ad based) model ... not a replacement - and can apply equally to native, RIA or browsers based apps.

JavaScript versus Objective-C
This just seems like a silly comparison, but I think the real versus here is AJAX/HTML5 against Objective-C. Which seems to unfairly leave out the 3,000 other options for building web apps.

HTML versus XML
See above about the web turning into a data center. Like the Freemium comparison, XML isn't a replacement for HTML (though some developers would argue that XHTML is...) - it's another format for storing data without storing UI. Want UI? Get HTML. Want to create your own UI? Get XML.

Basically, let's return to the graph and consider what's really going on - the web isn't disappearing into oblivion, it's just that there has never been the all consuming webtop to eliminate the rest of an ever growing list of neighbors. It also seems premature to put nails into the web's coffin before Google has managed to release the Chrome OS, the first serious webtop concept to be produced in something like 15 years (I know, I developed one 15 years ago).

Maybe I'm biased though. I'm reading my email in a native app which actually just creates WebKit instances, while debugging my work in the cloud and listening to music in AIR. Does the use of AIR apps, which is closely associated with Flash, which is a plugin for web browsers ... mean that the web is dead or very much an undeniable part of our daily lives now?

My old boss and I had an ongoing argument we both knew was relatively ridiculous ... PDA's versus phones. Ridiculous because each debate resolved to the same conclusion ... that they'd converge to the point where you couldn't hold the distinction. That seems to be the real story of the web - that eventually you won't know where your native desktop and web technologies divide.


Steve said...

For perspective, Wired magazine is the same magazine that was touting the magic of the new economy. Howe we'd have decades of low inflation unstoppable growth.

In a word, or five: Wired is full of shit.

Always have been, always will be. The web is not dead. As a proportion of traffic it doesn't dominate but, duh, it's largely text and text compresses REALLY well. Video, on the other hand is a giant bandwidth suck.

What would be a far more useful graph would be a study showing what people spend the largest proportion of their time doing. I guarantee you that web browsing would still dominate that.

So the real headline for their article is:

"The Web takes up less bandwidth than other stuff!"

Wow, stunner.

Steve said...

Oh and one more comment as I glanced back at that chart again. Email basically disappears. If I'm on gmail is that browsing the web or using email? If I'm on facebook sending somebody a message is that email or web?

Josh said...

Yeah, that's a good point. From the looks of the graph, I'd assume they make mail equal STMP/POP/IMAP ... maybe?

But you're right, a more useful graph would be comparing behaviors to toolsets, not just overall traffic. Especially since the trend in web development has been to reduce overhead heading to browsers. The more AJAX implementations, the less traffic since you don't have full page refreshes for every action anymore. A more nimble web is certainly not a dead one.