Despite my feelings about who will find TabletPCs to be really useful, I've decided to try using one as my main system. The TabletPC I'll be testing is a prototype of Acer's 700 MHz mobile Pentium III-based Travelmate 100. But the fact that it has built-in support for 802.11b-based wireless Ethernet--as all tablets should--means it may have a problem unique to portable devices with built-in WiFi.
More and more systems are beginning to ship with built-in WiFi. Considering the fact that I have routinely extolled the virtues of WiFi in this space, you'd think that's a good thing.
While 802.11b is a wireless networking standard, it's a no frills standard. It offers a baseline of wireless functionality without addressing needs such as management and security that more mature networking standards cover. Not that the IEEE (the standards-setting body that oversees WiFi) is sitting still. In addition to the 802.11b standard, there's an alphabet soup of improvements to the standard (including 11a, 11g, and 11i) that may eventually become as commonplace as 11b is today.
But, while we wait for that to happen, most of the WiFi vendors are more than willing to provide proprietary technologies that fill in the gaps. Perhaps the best known example is Cisco's wireless security technology called LEAP. 802.11b's biggest weakness is the lack of a reliable way to authenticate wireless users before they're allowed to achieve a wireless connection. The problem has spawned urban legends of how notebook and PDA users are able to walk up and down San Francisco's Market Street without ever losing their Internet connection. Presumably, they're surreptitiously roaming from one company's wireless access point to another.
That problem is due to be addressed with a technology called EAP (extensible authentication protocol) when the IEEE's Task Group I completes its work-- the result will be the 802.11i standard. Until then, companies that make wireless networking gear like Cisco's Aironet have an opportunity to close this hole with a proprietary technology that not only solves the problem, but locks out other wireless vendors' hardware from being deployed on the same network.
For example, if you've deployed Cisco's Secure Access Control Server (ACS) to authenticate users trying access the wireless network, you'll need Cisco-based wireless Ethernet adapters such as the Aironet 350 Series.
The solution works quite well, unless you want to buy a system like Acer's TabletPC-based Travelmate 100 or any of the umpteen notebooks, PDAs, or other mobile devices with built-in WiFi. Almost none of the companies that make systems with built-in WiFi make the WiFi components themselves. Like Acer, most manufacturers OEM that technology from a third-party WiFi provider. This poses a serious problem for both buyers and sellers of devices with built-in WiFi.
Suppose the device you want has built-in WiFi from a vendor other than the one that your company has standardized on. If your wireless infrastructure is enabled for a proprietary feature like LEAP and the WiFi guts of the TabletPC you want come from a company other than Cisco, you may have a problem. Even if you're aware of the issue before you decide which device to buy, you've got another challenge: how to find out which proprietary wireless technologies are supported by the devices that you're interested in. A random sampling of Web pages at Acer, Dell, HP, IBM, Sharp, and Toshiba reveals that this information is readily available only from Toshiba. The company's notebook offerings appear more flexible than others, supporting either of the two most often OEMed wireless technologies--Agere's Orinoco (recently acquired by Proxim) or Cisco's Aironet.
Wireless vendors like Cisco aren't the culprits. Not only did Cisco solve a problem that desperately needed solving, but there are plenty of other vendors of wireless Ethernet hardware that have built similar "walled gardens" with their proprietary technology--technology that you or your company could already be relying on.
In some cases, the problem can be overcome, but the solution is a kludge. For example, the Acer tablet that I'm testing can accommodate a different wireless networking adapter in one of its PC Card slots. But if you have to add your own wireless technology to a device, how do you justify the additional expense you've already incurred to have similar technology built-in?
Time will also help to solve the problem. As the IEEE standards for wireless Ethernet mature, proprietary solutions will fall out of favor and the risks associated with buying devices with built-in WiFi will diminish. But until then, buyer beware.
Have you encountered a WiFi compatibility problem? Let me know. Leave a message in our TalkBack forum, or e-mail me at firstname.lastname@example.org.