Friday, 18 January 2008

Nuances of Networking

I don't quite know what to make of this.

There's a nuance there that I am not getting. The article is expounding on research to create a new "world-wide-web" Now, thanks to the work of the WIP project, we may be on the brink of a new internet, a new world wide web. One where users can spontaneously create their own networks, in minutes, and with any kind of data device – mobile or fixed, handheld or deskbound. It means completely reinventing the internet, retooling its underlying technology, creating new operating principles and defining wholly new communications protocols so that it all works with any technology.

Huh? For what purpose? What benefit?

Right now the way the Internet works, I can spontaneously communicate, share photos, videos etc. with friends, family, strangers, etc. from any network enabled device. I already alternate between desktop, laptop and blackberry and use all the same programs. So exactly how is this Web 3.0 a new idea?

The examples provided don't help. I'm assuming these two were separate examples, not one and the same.

An example helps illustrate the concept. You live in an apartment building. You find neighbour’s wifi connections and invite them to join a new ‘building network’ with a few clicks. Now you can share and communicate with everyone.

And you can't do this now? This is how companies have been building their networks for years. How do you think employees "share and communicate" now?

You all have internet connections via an ISP, ranging from 1, 2 and 5 megabits/second (Mbits/s). You decide to pool your money and rent a fibre-optic line that handles voice, data and TV for the whole building. Suddenly you all have 10Mbit/s connections.

Again, you can do this now. Cost is a little prohibitive, 10 MB Internet ain't cheap. Also, you do have to "network" the devices, but it is doable with existing technologies.

Another scenario. You go to a gig with some friends, set up an ad hoc network, and you can all communicate via text, voice or image for the rest of the day, all for free.

Again you can do this now. There are IP enabled devices out there you can use for "free" networking.

What are the obstacles to doing the above examples now? Well technical expertise of the user is required in some way. Which the article explains away with this flippant statement "user-friendly set of technologies and standards that will mean any user, anywhere, can identify and network with any nearby devices. Without any technical expertise whatsoever"

Uhh, riiiigggghhhhhtttt. To use a technical device, you need some technical expertise. Using a keyboard and mouse are two of the most basic technical abilities that people need to know. Technical expertise is a matter of degree. So to what degree technical expertise is needed, is a very subjective opinion. I find setting up wireless networks easy, some other people, not so much.

Also, there's that massive bug-a-boo called "security". Forgive me for being cynical but "easy-to-use" and "secure-for-the-user" have always been at odds with each other. The easier a network is to "use", the less secure the data is on that network. To say I have my doubts that this group will solve that problem is a massive understatement.

This statement slays me, "But if separate values are used for identity and routing, then this isn’t a problem, even if the user is walking through a park. We’ve successfully separated the two functions."

We already have separate keys used for identity and routing, its called, oh I don't know, a user-id. My readers know me as "Catelli" You think I only use that ID from one computer? Also, imprinted on a network device, there already is a unique identifier as well, the MAC address. Unique identifiers are used all over the place. The application used is what determines what kind of and what combination of unique identifiers is needed.

What it comes down to, for reasons I fail to fathom, is that this project is trying to take the hidden functionality of networking, write it directly into the application layer of the network, and make it directly part of the user experience.

How this solves anything is way beyond me. I only see it creating a different set of problems. It appears to me that this is the Microsoft Bob project all over again.

No comments: