When traversing a wild unknown terrain that constantly changes, it's pretty obvious that you need a guide, or at least a map. The latter can take any form from a pencil sketch to a printed map to the latest newfangled GPS gizmo.
The Web is such a growing, throbbing, ever-changing terrain. Sure enough, one of the most important tools for it is a guide in the form of search engines that quickly transformed into full-fledged portals. From there they quickly blossomed and multiplied into veritable thousands of such engines worldwide, and mutated into all sorts of variants like private corporate portals as well as the likes of vertical portals.
As with the Internet at large, when it comes to portals new concepts, technologies and twists continue to come out of the portal woodwork on an almost blistering pace. For instance, the trend moved from spiders looking for keywords to engines basing website listings by popularity (as in the case of Google). Another example is from a search engine to "human" portals as with the case with About.com. Or from your classic search engine and portal to that of business-to-business (B2B) and vertical portals. And there are many others.
However, in all this innovation two things almost always remain constant:
- The fundamental function is to help users find what they are looking for, ideally but not limited to providing them with good search results;
- They are all based on a web site.
Now the latter constant may seem trivial, as everyone almost assumes this has been the way it's always been done and is thus the only way to go.
Until you consider how a simple change to that assumption can yield huge differences.
The goal remains the same: help people find what they want. But the inner workings, benefits and results can be extraordinary.
This paradigm shift is something that has been getting overused and overhyped in the press lately, but is something can have wide implications to portals. It is P2P, or the peer-to-peer paradigm. In P2P, instead of users like you connecting to a website to access pages or get files (music, etc.), you instead connect to other users in the network, possibly many thousands of others. It has been popularized by Napster, where, instead of getting music files from centralized websites, many of which are dead links and do not exist, one can instead connect to Napster's network of tens or even thousands of users (out of more than 25-30 million) connected at any one time, and find the songs you want from anyone of those connected that has it.
What's interesting is that this is a concept that can be applied to other fields. From research to sharing (files, documents, etc.) and many others. There are legal implications of course, but that is another story. Our focus is how such a technology can be developed to the realm of portals.
Instead of a portal being housed in one centralized location and accessed by many from this one location, with all its content also from one location, and whose links will connect you to other servers and websites as well, wouldn't it be revolutionary if you can instead also tap not just servers and web sites, but also the computers of thousands or even millions of computers of people like you who are connected to the network?
Security concerns are addressed in the form of security measures built into the peer software used to access the network, as well as the fact that people connected to such a network cannot just access computers with impunity, but access only the directories and files that people allow in their computers.
So if you are searching for a particular type of widget, say, instead of going to static portals or making a search from a search engine, wouldn't you like to also find the documents related to that criteria from more sources, namely thousands or even millions of other users? Such a software could also do the double duty of not only making a listing of successful matches to your search criteria, but also rank them effectively according to accuracy, perhaps like the accuracy of the likes of Google.
Here, instead of accessing directories from from a central location or making a search from the same place, you get to do it across all other users and their computers connected to the network at that moment, even perhaps also including the static web sites as well.
This raises a lot of questions like how such a technology can be effectively developed, let alone deployed. Yet there are already existing nascent systems that do this. Napster and Scour are a couple of the earliest examples, where users go to a centralized directory to find songs and related files from other users and sites, but when matches are made the file transfers are directly done between users without the aid of the directory. Pure examples of this peer-to-peer paradigm include Gnutella, Freenet and many others, who use a system that is purely distributed and do not even have a centralized directory.
These systems still have a long way to go as to robustness, but what makes it amazing is that they already exist. Now.
This even eclipses the other benefits, like how such a system's chances of ever crashing is slim, to name just one benefit.
And to deploy them in the realm of portals and others will only be a matter of time, just as it can possibly also be only a matter of time before such systems truly start improving to the point of being good enough for use by the mainstream of users.
Just think of how eBay might change if it were somehow crossed with the peer-to-peer concept. This could have a revolutionary impact on the concept of the online auction. That's just one example of the turbulent and far-reaching changes that P2P may unleash.
Nicholas Mercader is publisher of the Business Models Insider, a newsletter which explores new economy concepts that stretch the realm of the possible.