It is no secret that the wireless LAN industry has made a huge mess so far of the opportunity that is known as 802.11n. This standard, which holds the promise of application level, i.e. real, wireless throughput of over 100 Mbps has had a rough birthing so far. It has survived a contentious IEEE proposal process, which took a detour to be "helped" by an industry consortium that managed to hammer out an uneasy peace between warring camps.
The cease-fire agreement included the "Draft 1.0" of the standard that allowed manufacturers of consumer WLAN gear to spawn the current crop of "draft 802.11n" products. These half-baked, rushed-to-market products are buggy and interfere with legacy 802.11b/g WLANs. They also, with the rare exception, fail to crack the magic 100 Mbps mark under best-case conditions and mostly lack the gigabit Ethernet switches that should go hand-in-hand with products that prominently display speeds in the hundreds of Mbps on their product boxes.
As a result, this motley band of products has been met with near-universal disapproval by reviewers and analysts. And even consumers seem to have gotten the message causing disappointing sales that have driven manufacturers to rethink their strategies. So manufacturers have been spinning the bad reviews, retrenching and plotting new strategies to get sales moving on these high-profile products that are sucking up their marketing budgets.
Which brings us to yesterday's surprising announcement by the Wi-Fi Alliance that it will certify draft 802.11n products.
I am certainly sympathetic with the pressures that have been brought to bear on the Alliance, since there is a huge pile of money at stake in the 11n race. But, simply put, the Alliance blinked, and has started down the path of devaluing its vaunted "Wi-Fi CERTIFIED®" branding. Remember, this is the same organization that threatened manufacturers with loss of certification of any product claiming 802.11n capability not too long ago. And now it has not only changed its position, but completely rolled over.
Glenn Fleishman has compared the Alliance's announcement to its earlier and successful WPA program. In that effort, the Alliance was key in bringing about a real improvement over the flawed and compromised WEP encryption, months, if not years, before the IEEE standards process would have yielded. The Alliance's program took elements of the draft 802.11i standard, which was also having a long and difficult birthing, wrapped a marketing and branding campaign around them and included a certification process to ensure product interoperability.
At first glance, the new 802.11n "two phase" certification process looks like the same thing. So what's the problem?
In a nutshell, 802.11n is not ready for market, despite the claims to the contrary by chipmakers and consumer WLAN product vendors. 802.11n is perhaps the most complex WLAN technology project attempted to date. The issues are not easy and the standard is not "almost done". But those are just mere details to the WLAN industry spin machine.
Take this wonderful example of double-speak taken from a press release that Airgo sent around yesterday (emphasis mine):
For those of you not accustomed to reading these marvels of wordsmithing, what this says is: "you can believe us that Draft 2.0 is the real deal, but don't believe those weasels who told you that Draft 1.0 was the same thing". And this from the folks who have from the beginning been wringing their hands that their competitors were besmirching the good name of MIMO technology by applying it to anything under the sun that had more than one antenna (ok, maybe more than two).
If the rah-rah statements in yesterday's stories about the maturity of 11n seem familiar, it's because chipmakers said the same thing back in January about Draft 1.0. Of course, that was before the 12,000+ comments on that draft came back. But hey, everyone bought the story last time, so why not try it again?
No folks, we're a long way from Kansas and you should pay attention to that man behind the curtain. Draft 11n is still a science project and should never have been brought to market at this stage in its development. Didn't Atheros set enough of an example with its Super-G "bad neighbor" debacle? And you mean to tell me that all the MIMO rocket scientists, who are much smarter than I'll ever be, didn't know that 11n's 40MHz mode would interfere with 11b/g gear?
Of course they knew, but were just hoping we feeble-brained reviewers were too sloppy to give their products the real going-over they deserve. But all these problems are fixable with a simple firmware flash or drive update, aren't they? So where are those updates, folks, especially the ones to fix the "bad neighbor" problem? Maybe that problem hasn't been fixed because the chipmakers are quietly spinning new silicon to add the hardware changes required to support a real fix.
As I've said before, the consumer WLAN industry is not interested in solving real user problems. If it were, companies would drive down the cost and drive up the availability and variety of dual-band gear in the U.S. They are already producing the gear for Japan, where WLANs would be virtually impossible without it, so why not here? They would also dial way back on the hype, do better consumer education, stop coming up with whizzy marketing names to try to differentiate the same technologies and do a better job of getting the gear to stick once a consumer gets it home.
Until yesterday, the Wi-Fi Alliance was the only organization that had some ability to bring WLAN manufacturers to heel. But with its decision to certify non-standard products that provide no promise of upgradability to the released standard, it has embarked upon a different path. One that perhaps brings smiles to WLAN marketeers, but sends a confusing message about what the Wi-Fi mark stands for.