November 01, 2004, 3:23 PM — Just for fun, I recently went to my local CompUSA (which is the last of the big computer stores, even though it's starting to look like a Circuit City or a Best Buy with all the home entertainment gear, which is OK, I guess, because those guys sells a lot of computers and related stuff - but I digress) to look for .11b-only equipment. Mind you, not .11g, which includes backwards compatibility to .11b, but .11b-only. I guess I wasn't surprised when I couldn't find any such products in the store (although there are a couple on the CompUSA Web site). To be fair, I headed over to Staples afterwards and did find some .11b-only adapters and routers there, but the price differential between these and .11g was only about ten bucks. So, who would buy .11b only? It's similar to asking who would buy a 10 Mbps-only Ethernet card, when 10/100 products are almost free today.
Indeed, most of the .11b products I've seen on the Web lately seem to be bargain-basement blowouts, with adapters at around ten bucks and, believe it or not, routers at the same price after rebate! Clearly, no one is making any money here, so it would seem that .11b-only products are likely doomed. Will they disappear entirely? As it turns out, this is a more complex question than what I've presented so far might imply.
First of all, there's the issue of installed base. That there are millions of in-service .11b adapters most certainly doesn't imply that more will be sold, but the large number of APs and routers installed just might. Most public-access ("hot-spot") services are .11b-only, and many home users, even with broadband connections, really don't need more than .11b's 11 Mbps. I'd really have no problem in recommending bargain-basement .11b adapters to most residential users. Technology does indeed march on, but that doesn't necessarily obsolete the products in use today.
There's another, and perhaps more interesting issue, though. .11b-only products should cost less and consume less power than more complex, multi-technology offerings. Today's Wi-Fi PDAs (like the Dell Axim and Palm Tungten-C) and cell phones (like the Nokia 9500 and Motorola MPx) are in fact .11b-only, at least partially for these reasons. Of course, the announcement of highly-integrated, multi-standard Wi-Fi chipsets (like Atheros' single-chip a/b/g products; see) probably means that future client/subscriber products will be multi-standard. Indeed, Farpoint Group today recommends that enterprise end-users buy only a/b/g adapters so as to allow the maximum degree of flexibility and connectivity no matter what the situation. And, if (when?) .11b finally bites the dust, client devices will maintain all of the utility they have today. I have an early Centrino notebook (an IBM X40 see) that's .11b only. It's just itching for an upgrade and I'll replace the mini-PCI WLAN adapter at some point (well, when the warranty's up, anyway).
It seems very likely that .11b will eventually fade from the scene. .11g replaces it via backwards compatibility, but most users will prefer the throughput of .11g regardless. And I think most of us will eventually migrate to .11a because of the huge amount of bandwidth (24 non-overlapping channels, vs. 3 for .11b and .11g) available to this technology. The .11n standard, which will push per-channel performance above 100 Mbps (and perhaps to 500+ Mbps), will also likely provide backwards compatibility to .11a, but perhaps not to .11b or .11g. My guess is that we will eventually abandon the 2.4 GHz band where .11b and .11g run, and move entirely to the 5 GHz. of .11a and .11n.
