Archive of UserLand's first discussion group, started October 5, 1998.

Re: Weblogs and Search Engines

Author:Chuck Shotton
Posted:5/29/1999; 7:36:39 AM
Topic:Weblogs
Msg #:6870 (In response to 6841)
Prev/Next:6869 / 6871

The big Search companies have missed a big one. Weblogs and search engines will become the same thing in a couple of years. Think about it.

They didn't miss it. It's just a harder problem to solve than it first appears and most developers opt for the easy one-size-fits-all server-side solution that precludes meaningful use by individuals. I gave an ad hoc presentation on April 3, 1995 at the first WebEdge conference on this very topic, just as the first "portals" were coming on-line.

The discussion centered around the lack of common, distributed infrastructure afforded by the emerging commercial, server-centric Internet model. I spent about an hour on the whiteboard showing how the Web should/could have been put together had it been the product of more than a single individual (i.e., Tim Berners-Lee). It boiled down to one thing. There wasn't (and still isn't) a common data model with a peer-to-peer network to support it.

We are stuck in a universe of millions of clients and only thousands of servers, with vast amounts CPU power and bandwidth going unused. Specifically, there is very little upstream data from clients and virtually no CPU processing happening beyond rendering HTML pages.

As for data, XML has just shifted the problem from non-standard proprietary data formats to non-standard open formats. And we've seen overtaxed engineering staffs at portal sites consistently opt for the low tech, low difficulty, server-centric implementation of on-line applications every time. This continues to propogate non-standard data formats for use within each little commercial site's island of servers and no interaction with other servers or client applications.

What the Internet needs is a set of true peer-to-peer applications that run on the client side. Until that happens, the old client-server model will continue to be a barrier to any sort of meaningful distributed or decentralized computing on the Internet. There are a few companies trying to change the model, but I can tell you from personal experience, trying to change the infrastructure of the Internet without a killer app to drag it along is a lost cause at this point. There is just too much inertia caused by the massive investment in server-centric sites to see this move happen if it isn't through some sort of mass adoption at the grass roots level.

There are responses to this message:

This page was archived on 6/13/2001; 4:50:31 PM.

© Copyright 1998-2001 UserLand Software, Inc.