Slashdot Mirror
Being Wireless: Viral Telecommunications
sh4na writes "3G is out before it is ever in... because, as Nicholas Negroponte puts it, the *real* next generation is the Wi-Fi "lily pads and frogs" concept. Wouldn't it be great?"
← Back to Stories (view on slashdot.org)
He mises one crucial point, the backbone, this posting for instance isn't going to magically bounce from the middle of the UK to Exodus(?) by 802.11 alone, as illustrated by the recent crackdownby cableco's on publicly listed access points they're reluctant to support an essentially public network that runs contrary to their business model, and transcending the traditional backbone requires organisation and capital, something absent from P2P systems.
A concern is the finite amount of spectrum available in regards to the scalability of P2P wireless systems, as the number of users increases so does the baseload just to maintain the system, some clever managed routing will be required along with a wired backbone between nodes, if you use daisy chain style off-air repeating between nodes you quickly deplete spectrum and diminish the benefit of local frequency replication, basically the "everyone shouting in a crowded room" scenario.
"Distance decay" is a feature of the traditional phone network yet on the net people no longer communicate on the basis of geography, did that Wired article come from a server in Silicon Valley, New York, London? Does it matter, I don't particularly care, I'm just interested in the content. However the "lily pads and frogs" architecture is deeply tied to locality, it's easy to communicate with local nodes but it progressively gets more difficult the farther you go, again this leads us back to the backbone problem.
Another issue is misuse, free wireless reminds me of the net of yester year, you could for instance use SMTP servers all over the globe and the vast majority of users didn't abuse that facility, but obviously the small majority of spammers swiftly made that a thing of the past and continue to annoy us today, how would open wireless networks be any different? Control is needed, which leads back to structure and capital.
Call me a pessimist, but it's not quite as rosy as he makes out.
Any bets on how long this article will be available until the guy retracts it claiming "Sorry... It sounded like a good idea but do to the overwhelming number of e-mails I've been receiving from Slashdot users telling me what an idiot I am, I am going to retract my article."
... the list or problems not addressed by his article is huge ...
There are numerous flaws with his article, all of which have already been pointed out by other posters so I wont bother restating them. But here's a summary.
1. Data access. You are responsible for the data on your router. If your neighbor sets up a kiddie porn site after hijacking your net connection, at the very least your connection will be shut down and your life will be highly inconvenienced. At most, you go to jail.
2. Roaming. Switching from one wireless network to another isn't this magical thing that just happens transparently and seamlessly like this article suggests. If you exceed the range of one wireless network in the middle of a download, your download isn't going to magically start up again when you hit the next wireless network. (Granted you can "resume" if you are using proper software.)
3. Money. If dozens of people start cancelling their cable/DSL because they are all taking part in a neighborhood wireless Internet pool, the cable/DSL guys are going to raise prices or go out of business.
4. VoIP. Hahahahaha. Enough said.
The biggest problem that the article barely touches on is that there aren't a huge number of non-overlapping channels. If three homes in a row all use the same channel, stand out on the street and watch what happens to your signal quality.
As others have noted, if one home is on xDSL from company Y using one public IP address and the next "hop" is using Cable from comany Z using another public IP address, at best case scenario, your data transmission will suffer a temporary silence. In more likely scenario, you lose connection with an associated AP, your PC attempts to renegotiate with an AP, grabs IP information and reinitiates IP connection.
How long does this take? Too damn long for VoIP or even a web page to load. You could of course set IP Leases to expire every second to help, but talk about broadcast storm.
Also, until full T1's or T3's start getting run to every home, DSL's and/or Cable modems just can't handle 25+ people all downloading files acceptably. Put the number up around 100 and it is not a pretty picture to paint.
Even if the performance were to be acceptable, how long do you think it would take for DSL providers to realize that for every 1 customer, 15 are using it without additional revenues for them? Expect heavy handed TOS to put and end to that quick.
I think the idea is right, but his visualized implementation is flawed. Now if he said a meshed network powered by WISP's with a cooperative agreement, that would make more sense to me.
Just my $0.02 worth.
I case you don't know him, he is the perpetual tech optimist Gilder Tech Report.
These wall street, talking heads are always so optimistic about technologies that they fail to see their shortcommings. I love new technology as much as any slashdotter, but do you think for a minute that reliable wireless data will be built on a technology that can be knocked out with a 2.4 GHz cordless telephone?
Gimme a break.
-ted