Small Fish, Big Pond

Like I said yesterday, Wired’s article is already making waves. Chris Anderson was interviewed on NPR about it this morning and this afternoon it made the news crawl on CNN.

One thing I like from the NPR interview is that Chris mentioned that by dead he’s talking about Web transitioning to Mobile. Which in a way is very true. Although he still talks about how applications rule and that they will kill the web.
Here’s an experiment to see if he’s right: Use only apps, no web browser.

Go 2 days without ever opening Firefox, IE, or safari, chrome, etc.
Don’t use Google (it’s a WEB page).
Try getting the things you want done with only dedicated web apps. No diversity of the millions of online web pages, just the 20 or so apps you can load before your phone fills up.
Don’t be fooled by apps that redirect you to a browser, they’re cheating.

Basically Chris’s prediction of the future of the web is where the multiverse of web pages is boiled down to a handful of corporate apps that port and filter the web for you. Much like the archaic AOL days in internet prehistory. And that scares the shit out of me.

Luckily he’s wrong!

Rob Beschizza edited the fact distorting graph used by Chris for the wired article to better fit reality. Pay close attention to the red “web traffic” That is “dying”. This is just the same graph but adjusted using the same data used for Wired’s article to reflect the actual amount of traffic passed in each category.

In Wired’s article it shows web use as a percentage against other high bandwidth internet traffic. Now that we can see the actual amount of web traffic we can see that in the last 5 years the web has almost tripled. Rob summed up Cisco’s data best:

Assuming that this crudely renormalized graph is at all accurate, it doesn’t even seem to be the case that the web’s ongoing growth has slowed. It’s rather been joined by even more explosive growth in file-sharing and video, which is often embedded in the web in any case.

In regards to using “bandwidth” to measure the value of internet traffic.

Does 50MB of YouTube kitteh represent more meaningful growth than a 5MB Wired feature?

It’s worth noting that we’re talking generalized numbers and graphs and that there will be a bit of variation in the data. But the web is still a LONG way from dying. Harry McCracken at Technologizer has another great article pointing out other technologies that have “died” recently (Hint Facebook died 2 years ago but Vinyl is alive and well).

A few people were down on Cisco for promising the “Next Generation” of the internet and then just releasing a new router a few days ago.

I’ll skip over the fact that EVERY manufacturer claims their gear is Revolutionary, Game Changing, or the Next Generation; I mean the iPad is just a big iPhone but apparently it’s “Magic”. For me magic is then a beautiful girl comes out of a genie bottle, calls me “Master”, crosses her arms and bobs her head, and creates a huge feast out of thin air.

So Cisco didn’t just create the next generation of the internet. But they built the device that can handle the throughput for the next generation of the internet and that’s just as important. Bullet trains may be the “next generation” of rail travel but without rails it’s just an expensive, immobile, hunk of metal. Cisco makes up about 86% of the internet routing devices, so when they see big bumps that means all of the web benefits.

It’s easy for casual home internet users not to realize how important the backend of the internet is but that’s only because the internet has never run out of bandwidth. Can you imagine what it would be like if your home DSL connection only ran at dial up speeds from 2pm-8pm because the net was overloaded with Hulu streams?

Luckily backend technologies are keeping well ahead of current demand and this is the moment when potential expanded to three times its size. Next Generation it may not be but this is still quite and accomplishment.

The CRS system devices are powerful on their own but their big claim to fame when originally developed was the ability to cluster the devices to create one super router. Through this clustering a single location can have the theoretical routing throughput of 322 Terabits of information. To put that in perspective as Cisco states, 322 Tbps is equivalent to transfer the entire printed collection of the Library of Congress in just over one second; every man, woman and child in China to make a video call, simultaneously; and every motion picture ever created to be streamed in less than four minutes.

Already Cisco and AT&T (ironically both my last 2 employers) are researching to put the new tech to use to create “thicker” backbones. AT&T owns most of the backbone connections that link regional carriers and even most of the undersea links coming into and out of the US. Unlike the latest smartphone, tablet, or laptop release event, or any of the new up and coming websites at SXSWi this development by Cisco will directly impact you and you life.