What the hell is a ‘proximity server’, and why should you care?
Bridging last-yard data
How interesting: more strange fusions at the edge of the network.
CeBIT saw the emergence of the personal wireless gateway. As Guy Kewney reported here and as we discussed further, here, the "phone hub" or PMG "personal mobile gateway" allows you to move your mobile phone - or other wireless-enabled trinkets - between networks. For example, Siemens PMG-enabled phone uses your lower-cost landline when you're indoors, and when you're out of the house, it finds the GSM/GPRS connection.
It's a device category that needs a snappy name, but for now PMG must do.
Now meet its cousin: the proximity server, or wireless access point. Both are indeed wireless gateways.
The Koreans already have tens of thousands in use. These local points allow you to download software or updates onto your phone while you're in a Radio Shack. The driver isn't just software, of course, but a bigger prize: M-Commerce. "Shop" buttons are increasingly common on Asian phones and a Bluetooth-enabled handset will allow the vendor to push special offers at you.
This netherworld - you're in a supermarket, but looking at the screen of a computer rather than the packaging on the shelves - has some interesting implications. If the on-screen menu takes your eye, rather than the visual signals emitted by the packaging, then we can see a world much like the one Alex Cox envisaged in Repo Man - where every object on the shelves has identical industrial packaging, differentiated only by name. "Beans", or "Toilet Roll".
Modern marketing and the advertising 'industry' began when the shopping experience was de-personalized, and we no longer had a specialist shopkeeper to recommend his wares. It's a far cry to suggest that standing in a sterile supermarket looking at a sterile computer screen for what to consume constitutes a "repersonalization" of shopping - it sounds like an emotional Siberia, to me - but it does have implications for the world of branding and graphics design.
But we're getting a little ahead of ourselves, here.
These "Proximity Servers", as we've heard them called, are already here. WideRay stations are a familiar feature at trade shows, dispatching the show agenda and exhibitor details to Palm and PocketPC PDAs.
The difference is that in the coming months is the clients capable of receiving wireless data will be arriving in significant volume. By the end of the year, WideRay CEO Saul Kato told us yesterday, all application-capable phones will sport a Bluetooth or InfraRed port. And he meant the US, rather than Europe.
Kato is based in San Francisco and has a lot of interesting observations to share. The WideRay is a gateway: the vendor calls it a "wireless service point".
At the back end it takes GSM/GPRS or Ethernet in, has room for cached data and is capable InfraRed or Bluetooth or 802.11 out. (It's really a small Linux box, and you can add your own software to extend the platform).
But why default to GSM/GPRS, when 802.11 is "becoming ubiquitous". The answer is mostly cost, says Kato.
It's far cheaper to kit out hundreds of stores with SIM-equipped "Jacks", as WideRay calls them, because they make no assumptions on the existing infrastructure, except being able to pick up a GSM signal. You don't need to have DSL. The Jack doesn't even need a power chord: it's battery operated.
There's another advantage, however, other than volume. As anyone who has been to a busy trade show can tell you, lots of 802.11 networks in a small space do not make for a happy experience. A CeBIT or a CTIA show floor has a passing resemblance to a mall.
Making Money From Wireless
Like most in the industry, Kato doesn't see many people making money from WiFi "except at airports, and maybe coffee shops".
The vociferous clamor for 802.11 from the "laptop nation" hasn't resulted in a viable business model. People expect it for free, too, which makes VCs think "very interesting, but I'll invest my money somewhere where the public want to give my client some money".
As for the carrier side, what's in it for them?
Simple, says Kato. They're not really interested in delivering "mobile data" as making money, too. (We seem to have forgotten that money makes the world go round, or at least, must factor into the equation somewhere when VCs are asked to make capital investment decisions).
"ARPU [revenue per subscriber] is not a direct linear function of data," he explains.
"When you think about it, they'd rather you didn't use the network." Well, this is true of all flat-fee models. But of course, they do want people to make money off the services. And if a wireless service point in every store or location gives a chain an advantage, then it will do it.
The total cost of ownership of putting a "proximity server" in a store is about one-tenth of the cost of putting in WiFi. And stores have a much wider demographic reach than by appealing to members of "laptop nation" who happen to have their laptops with them as they wander by.
Kato hesitates to call this a Personal Area Network. While it takes connectivity down to the "last yard", he says it's more connection or transaction oriented. No pairing is necessary. WideRay writes client software for Symbian, PalmOS and PocketPC, which handles the exchange.
Another interesting aspect could arise if people exploit the cacheing aspect. Imagine a community space where you leave your photos, or a message board. Phone wireless is opening up a lot of social exchanges that we didn't expect before. In a cafe or a club you can leave a message on the board: "you with the leather pants - you cute." A crude example, I admit, but flirting was the driver behind text messaging, and that's become the most popular software in the world.
Kato is optimistic that the US won't be left behind this time. Despite the fact that the West Coast seems to be so dazzled by WiFi, that it's blind to all other forms of wireless, I'm sure this will become ubiquitous without most of us even realizing it. ®
Sponsored: Benefits from the lessons learned in HPC