Five chipmakers have certified Wi-Fi Direct reference designs: Wi-Fi Direct is a terrific addition to wireless networking where a device that offers a service can broadcast that service's availability, like printing or file exchange or what have you. It's a form of peer-to-peer networking that doesn't require an access point to intermediate, and is ideal for mobile devices, and devices that lack much of an interface. The first five reference designs have been certified a few months later than the original rough target announced last year. (See "Wi-Fi Alliance Peers into the Future with Ad Hoc Replacement," 13 October 2009.)
Wi-Fi Direct has a few things in common with newer Bluetooth devices that pair with less effort than in the original Bluetooth schema, and in that Bluetooth and Wi-Fi Direct both advertise available services. But the notion is that you get the speed (up to 802.11n) and security (WPA2 mandatory) of Wi-Fi with enormously simpler setup than connecting to a new Wi-Fi network for a moment and then setting up a connection with a specific device. And in cases in which you don't have an access point, such as trying to exchange a file between two mobile devices, it's extremely irritating. (On an iOS device, both parties could have a package like GoodReader that has built in WebDAV client and server software with Bonjour discovery, but you still need an access point to which both devices are connected, and security is an overlay.)
This announcement went around the world like a shot, but was typically covered incorrectly or incompletely in four ways. First, this is nothing new. The spec was announced a year ago; this is the culmination in silicon of that effort. It's great to see this implemented, because now we can move forward to have devices that support it.
Second, it's not yet available. The five certified devices are reference designs that other companies (OEMs or original equipment manufacturers), like Linksys, D-Link, Dell, Acer, and the like will build into products or relabel to sell under their own names. That means there's still some time to the market.
Third, this is host-side stuff—things to make a computer act as a Wi-Fi Direct enabler. It's not the technology needed for embedded client-side support in, say, an HP multi-function printer.
Fourth, there is no announced operating system support yet, even though Microsoft and Apple sit no the Wi-Fi Alliance board. That is not unusual for newly released hardware implementations of standards from the Wi-Fi Alliance or other groups. Apple and Microsoft both have near-term releases of operating systems upgrades on the timeline (Mac OS X 10.7 and Windows 8). It's most likely Wi-Fi Direct would appear in a new system, and might not be available in an older device.
Finally, and this wasn't addressed in any of the coverage I saw, you're going to need to see widespread adoption in mobile operating system platforms to make Wi-Fi Direct truly useful, and integration at a fundamental level of the OS. That means Android, BlackBerry, iOS, Palm WebOS, and Symbian (whichever version), as well as featurephone platforms from Nokia and others.
The reason is that mobile OS's, even the supposedly open Android platform, need to put Wi-Fi Direct hooks down into the driver level so that third-party developers can hook into a system-wide printing library that works with Wi-Fi Direct, or file-transfer support within apps.
Wi-Fi Direct is terrific, and I will be glad when it's widely available, But my prediction is that it won't have widespread impact until 2012. On the Wi-Fi timeline, that's perfectly fine. Each 802.11 standard as certified by the Wi-Fi Alliance has taken 2 to 4 years to percolate into the market. WPA2, rolled in 2004, is just now becoming the de facto security method, for instance. Wi-Fi Direct's greatest impact is on the future, not the present.
In this age of internet derived technologies, why is anyone expecting a hardware driven technology to gain any kind of OS or application support?
The last major hardware driven networking standard, Bluetooth, has been a protocol nightmare precisely because hardware folks defined the protocols. In contrast, Wi-Fi has been successful because the hardware people had a simple goal -- wireless ethernet. The internet runs on top of their efforts. Comparing the relative success of the two technologies says it all: Wi-Fi is broadly used in a wide variety of applications. Bluetooth is widely deployed for a very narrow range of uses. (Bluetooth goals were much grander than they have been realized.)
Any point to point technology faces 2 challenges: deployment and then use. What is the pragmatic use case and will folks work to change their patterns to use it? Finding a printer? Puhleaze... There are already at least two underutilized printer discovery protocols on Wi-Fi: UPnP and Bonjour. These respective technologies are driven by two of the largest OS vendors on the planet. Yet there is very little uptake. If the use case is file exchange, then how is Wi-Fi Direct better than USB sticks or native file sharing or email attachments or cloud data sharing (read DropBox and their ilk)? My point is simple: there are many ways to technologically skin these point-to-point data transfer cats. Does anyone need to use Wi-Fi Direct as their skinning knife? Hell, the fact that everyone today carries USB sticks is a testament to why Wi-Fi Direct is not going to solve this problem. Even if it is easy to use, it isn't easier to use than copying data on to a stick and handing that stick to your colleague. Or, if you are already using a cloud device (read iPad), transferring data only happens through the cloud. Wi-Fi Direct is moot here too.
Why do these Wi-Fi people bother? There's a good reason why the major OS vendors aren't embracing Wi-Fi Direct. There are better solutions and their businesses are driving different goals.
Andrew
P.S. Disclosure: I am a contributor to the UPnP v1.1 spec.