January 28, 2008 8:54 AM
Internet Reinforces Local Bonds
So often when you talk about the benefits of the Internet the conversation focuses on its ability to break down the barriers of distance, to unite disparate friends and relatives, to provide access to resources not available in your local community.
But there seems to be a growing understanding that the ultimate benefits that the Internet brings to society may be found not in tying together things that are far apart but instead enhancing the relationships that exist locally.
For example, this Wired essay entitled "How Email Brings You Closer to Guy in the Next Cubicle" explores how while broadband should enable anyone to do anything from anywhere that what it's actually doing is putting a premium on living in concentrated areas. The reason for this is the efficiencies the Internet can bring to day-to-day communications with the people you interact with regularly.
I can personally attest that this is the case for me. While I do email a lot with my parents and friends back in Minnesota as well as with colleagues across the country and world, when I look back over my inbox I have far more emails sent to and from people who live within a few miles of me.
But this capacity of broadband to unite communities is perhaps best highlighted by one of the most stunning facts I've heard in a long time.
In catching up last week with Matt Wenger, CEO of Packetfront America--Packetfront enables the creation of open access, multi-service networks--we got to talking about Vasteras, Sweden, arguably the fullest realization of what an open access, multi-service network can be.
The fact that blew me away was about them and how the deployment of a fiber network impacted their use of broadband.
Before this community fiber network was put in place, more than 80% of the traffic on local networks was outbound, pulling in and sending out information over the world wide web.
After the fiber network came into being? That ratio basically flipped as now more than 80% of the bandwidth being consumed is for moving data around within the Vasteras network, so neighbors talking to neighbors rather than users pulling in data from all over the Internet.
It should be noted that just because the percentage dropped, doesn't mean people on that network are consuming outlying Internet content less. Instead, it's a sign of just how massively demand for bandwidth in-network has grown, literally more than a thousandfold.
This trend is totally and utterly remarkable to me.
Basically everything on the Internet to date has taken a server-in-the-sky mentality where you're almost constantly sending and receiving data out over the world wide web.
To think that that paradigm has now fundamentally shifted in communities like Vasteras is one of most under-discussed potential outcomes of deploying a fiber network, and one that demands further exploration.
But never fear! While I don't yet have the answers to what's driving all that demand for bandwidth, my curiosity has officially been piqued and I'm on the hunt for more information.
Until then I'll leave you with this thought: the deployment of fiber networks can and should be considered the best opportunity we've had in a long time to not just hook people up to the global economy but also to reinvigorate the ties between people in their local communities.
Anyone deploying fiber, big or small, public or private, should be keeping this thought in mind, as otherwise we may end up missing out on one of the great unsung benefits of what broadband and the Internet enables us to realize.
Comments (4)
Geoff,
Thanks for a really thought-provoking piece....
I'd be very interested in what you find out! Please post as you discover more.
Some questions I'd (selfishly) like to have asked that would help me think about things:
1) Does the Vasteras network have large in-system bandwidth? That is, is it organized as an intranet--like large corporations or universities (or Lafayette, La!) where at least 100 megs internal is standard, regardless of inbound or outbound throttling. This would make 2 and 3 below more valuable to end-users than if it didn't have this policy.
2) Can you get data on how much of that traffic is P2P and locally cached in user machines? (P2P protocols typically "prefer" local IP addresses, I understand.)
3) Does the Vasteras network provide a large in-system cache that could be accounting for a proportion of this? (Most big ISPs do so.) There'd be a pretty big incentive for local networks to do this.
3) Any chance there are applications/local portals/etc. that have made use of the better bandwidth. (A 'killer app' :-))that is driving usage? Video chat/video phones come to mind.
4) You mention the sensible point that traffic to exterior sites probably hasn't fallen. That network cost is probably rising as the broadband capacity offered rises. But this phenomena would mitigate that. Any idea of how much?
5) How does packet-front explain this phenomena? (This would be a great tool in selling local networks a company like packet-front, it would minimize an uncontrolled cost of interconnection.)
The thinking behind these questions: Local provision of last mile fiber connectivity at large bandwidths is often considered risky--providers worry that if you give people big bandwidth they will find a way to use it. (Good assumption.) In truth, you hope they will...but the fee that you pay to the backbone provider significantly constrains what some providers are willing to offer. So we get the spectacle of some local fiber providers not offering the speeds that they are capable of for fear that a few users will use that capacity so eagerly that they will have to be provided for at a loss...raising the costs for other less eager users and threatening the competitiveness of those tiers of usage. The more local a network is the more this is an issue.
The possibility that there is some dynamic that results from big intranet bandwidth (regardless of outsystem, internet throttling)that would contain most of that new cost of big bandwidth offerings inside the local net where marginal bandwidth costs are negligible would be a big boost for hopes of local or regional ownership of last mile fiber.
Explanation via P2P dynamics would be a dynamic that didn't involve the local community doing anything.
Explanation via caching would require some very, very small investment...and could actually integrate with P2P if some clever things were done to cache requested material to a local P2P server that restricts connections to in-network clients.
Explanation via killer apps being some form of local usage would be fascinating to everyone. (Not just you.)
The real explanation is likely to involve all these things...and things I've not thought of...but the phenomena is VERY suggestive.
Thanks again...
Posted by John on January 28, 2008 6:40 PM
In San Francisco, the Meraki.net folks are establishing wireless mesh network infrastructure. Haven't seen anything about this topic, but might be worth fleshing out and looking closer. Suppose the survey of users produced similar indications. Would this then lead small rural communities to begin establishing low-cost community wireless mesh networks that acted as community intranets?
Posted by Tom Poe on January 29, 2008 10:28 AM
I've long felt that, particularly for those areas incumbents avoid because they can get bigger return on investment elsewhere, that robust local networks could be valuable. To my mind, the post and John's questions and comments articulate the issues well.
Posted by Bob Babione on January 29, 2008 1:54 PM
Very interesting piece, I'm apart of a community wireless network in Australia and due to some regulatory constraints here, the focus of our network has been much to do about local content and communications, rather than Internet sharing.
Despite the lack of formalized internet provision and members already having DSL Internet access. Our group continues to grow, fueled by people wanting to communicate and host content locally more than over the Internet.
Posted by Robert Hard on March 5, 2008 10:01 PM