Comments on: Web 2.0: Obsolete within three years?

I'm not talking about Joe Schmo. This is Steve Perlman, and he's not all that convinced that the state of the art has staying power.

[rcrusoe]

"... obsolete in three or four years." Wouldn't that be true of any of today's technology? I built my first intranet web server in 1993 (to distribute manufacturing drawings, etc.) and had a tough time getting my corporate IT department to give me a block of IP addresses for my company's locations. They kept saying "What are you doing? PC's don't use TCP/IP". (About four years later they launched a corporate wide Intranet project.)

Things change - quickly.

[pfletcher]

The critical part that needs to change for Web 3.0 or whatever is the need for symmetrical connections - so what if my RoadRunner can d/l at 6Mbs - it can only upload at 512Kbs - so much for interactivity.....

[Source00]

What a disappointing explaination... yes the web 2.0 stuff out there will be obsolete (or at least no longer buzzworthy enough for a trade show) simply because of the nature of technology. That's not insight... it's common sense at this point.

I was very disappointed in Perlman's reason for his disappointment with the show... no video in their front ends?... needing more use of AJAX like technologies... c'mon... give me a break. Can we actually talk tech and get some meat here?

My predition for what will make web 2.0 look like static HTML does to us now... will be the emergence of semantic technologies applying/automating contextual relevance (some journalist is going to "offically" brand it Web 3.0 so we might as well just call it that...), applications which enable and leverage personal data warehousing/data marts, AI enabled search, applicaiton of GIS and location based technologies (everyone seems to forget about that part), personal platform integration (your computer network knows your computers, your car, your phones... and whatever else integrates in the future... like your fridge, securtiy systems, HVAC systems, etc.)...

...so basically these apps that know and apply where you are (at most times), most everything you do, your likes, dislikes, secrets, habbits and latent tendencies... and then connect you to relevant people, content, systems, services, advertisers, etc... you know... help you live your life like a personal assistant would, shape your environment for you and remind you/inform you of various details. As you walk in your house, your network knows it is you and turns off the security system, turns on the lights or opens your curtins, turns on the AC/heat, reminds you to buy some lactose free milk and Splenda sweetened chocolate because you're almost out (actually it could automatically buy them for you via your preferred on-line enabled grocery store but it knows you want the illusion of control... which is just pushing the "Buy now" button on your fridg's UI), changes your buddy status on AIM, and turns of the TV to the Logo network because it knows what really makes you happy and that your wife/husband and kids aren't home.

That might be seem like a scary future but it's already started. Many of these technologies already exist... it's just a matter of application and integration... But no worries, freedom will ring whereas you should always be able to opt out... that is, unless you are using Google in any way and they'll just gather, profile and apply all your data anyway.

[StevePerlman]

As Charles mentioned, I made the obsolescence comment without qualification at the show, and then we caught up at the Rearden Web Crawl party last night where it was hard to talk. He did a good distillation of what came to my mind at the party, but of course, there is a lot to say.

First, I generally agree with the prior comments that you can count on Web technology obsolesence in 3 years no matter what, and that there is far more than video that we'll be seeing that wil extend what we are seeing today. But what I was really driving toward was that bandwidth and compute power growth are going to dramatically improve what we currently consider to be the "least common denominator" platform that we currently target Web applications toward.

When Netscape Navigator was introduced, there was 10s of thousands of bits/second of bandwidth (5 digits), and most of the Web was about text and small images. A T1 line, with 1.5Mbps (7 digits) of bandwidth at an office was considered the "gold standard" for connectivity, but was not generally available to the consumer (or many businesses). With the advent of early DSL and cable modems we saw 128kbps and higher (6 digits), and in the last few years we've seen speeds gradually grow to millions of bps (7 digits), catching up to and exceeding the T1 download speed. And very recently 10s of millions of bits per second (8 digits), have been available. So, now a significant percentage of your target audience now has a faster (downstream) connection than what used to be the gold standard T1 line, when Netscape designed Navigator (which ultimately shaped HTTP). When you combine that speed with the compute power to render decent video and animation (e.g. using Flash 9/Silverlight clients), it changes your least common denominator that you can design to. And, connections and processing power are only going to get faster.

But, very few Web2.0 applications that I've been seeing have fully exploited the richness that is possible in the post-T1 line era. A lot of the people Rearden has been working with are still wrapping their heads around the possibilities, and are often surprised to see what is possible with high-production value Flash/Silverlight and video, and what a wide audience is now reachable with the expectation of high download bandwidth.

And, given what we've been seeing people doing who have been working with us on advanced Web2.0 systems, when I compare what is in the labs vs. what they are showing publicly today, it is just night and day. "Life past the T1" will result in a sea change of the look, feel and capability of Web2.0 applications (no matter how you define Web2.0...and there are many definitions). You don't realize how many assumptions are tied around the limitations of a narrow pipe until you start working with the expectation of a broad pipe. And the stuff coming out is simply mind-blowing.

Steve Perlman
Presiden & CEO
Rearden