Feedback: Getting it Wrong and Making it Right
Honest inputs from sharp-eyed readers and responsible vendors
help me keep my stories straight
by Lee H. Goldberg
Editor's Note:
Since the original publication of this editorial, analogZONE has received a response from IDT to some of the criticisms leveled at it below by a reader we have identified as "Anon.":) In the interest of fairness, we've provided IDT "equal time" to address these charges and to explain their position.)
It's funny how a little feedback can motivate a person. As engineers, most of us are used to long hours and repeated bouts with disaster on the way to bringing a project to life. And while the cash bonuses or stock options (on the rare occasions when they're actually worth something) are certainly motivators, I've found most of us have the day-to-day encouragement to go out and slay the "dragon de jour" in small technical victories and the acknowledgement of the folks we respect. While the kind of feedback has changed a bit since I traded my soldering iron (does anybody use them anymore?) for a keyboard, the conversations I have with my readers is one of the things that keeps me sharp and makes me glad to turn the computer back on each Monday morning.
After broadcasting to what appeared to be a silent majority during my first months at analogZONE, the feedback from readers and manufacturers has grown slowly but steadily and is usually the first thing I read I the morning while I wait for my brain to come up to speed. Oddly, I've found that I value the critical notes calling me to task for an oversight or challenging my conclusions as much, or more, than those praising my work. One of the more interesting examples of a sharp-eyed reader providing valuable feedback came last week after I'd reviewed a new network search engine (ie gussied-up CAM) from IDT. Since it contained some interesting and credible claims, and good counterpoints to what I'd written, I thought I'd share it with you:
"Just wanted to comment on the networkZONE product announcement of the week for 9/13 that Lee wrote up about IDT. The product that he and IDT is talking about is merely a regurgitation of the specs for a custom CAM-based device that Cisco (the largest consumer of NSEs in the world by a whopping margin) has contracted all the major CAM vendors to produce since 2002. The major CAM players at Cisco are also bringing this product to market in 2005, some even with more powerful IPv6 enabling features. This is more hype than substance, basically PR to boost the perception of their struggling commercial NSE portfolio (a distant 3rd in revenue to Netlogic and Cypress).
"All the breakthrough features IDT crows about were defined, and in some cases patent-protected, by Cisco (except for ECC, which is just a few extra bits per entry added to Cisco's defined background parity scan). IDT's pre-announcement of this device has privately drawn a lot of fire from Cisco for potentially violating market-privilege clauses in the contracts that all the CAM vendors have signed with Cisco, and there will be likely consequences to IDT's market share at Cisco.
"Please forward to Lee for his information. Just wanted to let you know the real story behind this product so that your Editorial doesn't get suckered into giving credit where credit isn't due.
"Regards,
Anon."
Wow. I'm not sure where to begin here. I guess the first thing to do is to thank Mr. "Anon" for sharing some important insights that I might not be privy to as a journalist. While I know that Cisco can pretty much dictate its demands to IC manufacturers, it's enlightening to get a few juicy details of how the process actually works.
While some companies do give me limited access to such this sort of intelligence (as background under strict non-disclosure agreements), it's always helpful when a reader checks in and helps me understand a little more about what's going on with "private label" ASICs or other areas normally off-limits to the press. Besides correcting something I might have gotten wrong in a story, information like this helps me understand what equipment manufacturers consider important, what's possible, and what's eventually going to be available on the commercial market.
From time-to-time, I'll try to share some of the other feedback I get when I feel it adds to a debate, clarifies an issue, or corrects something I may have gotten wrong. As always, I'll be careful to make sure I'm not violating any confidences, NDAs (written or verbal), or exposing any legally-protected information. And while it's not always easy, I'll also do my best to make sure I'm not being manipulated to serve as a "stovepipe" for rumors or dis-information.
Happily, I've also got a good enough relationship with some manufacturers that they are comfortable enough to take me to task when they feel I've gotten something wrong. In fact I got one of those "real-time updates" last week when a product line manager at Atheros took issue with some of the things I'd said about their proposed version of the 802.11n standard for high-speed Wi-Fi transmission in my recent "WAN Wars" editorial.
One of his main objections to my analysis of the TGnSync group's proposal was my opinion that their use of a 40-MHz channel (vs the 20 MHz channels used currently) that their standard uses as a baseline made less efficient use of scarce bandwidth, and was not legal in Europe. In a nutshell, Atheros says that the TGnSync's higher transmission rates mean that a transmitter would occupy much less time on the air, allowing for more terminals to share the double-wide channels, thus providing more actual traffic capacity. Atheros also said that the TGnSync's 40-MHz channel scheme made better use of the spectrum by transmitting on a single 40-MHz channel that makes use of the guardband between the two narrowband channels: something that the WWiSE proposal does not do in its 40-MHz mode.
They also raised a few legalistic objections to the WWiSE party line which I'll relate here for the sake of completeness. They pointed out their interpretation of the ETSI regulations that govern the unlicensed band in Europe and, according to them, the current statutes do permit 40-MHz channel widths -- contrary to what WWiSE asserts. Atheros also expressed concern that their opposition may not be completely forthcoming about its position on the licensing of all 802.11n-related IP at no cost. They point out that TGnSync's "RAND-Z" ("reasonable and non-discriminatory, zero-fee") terms are only applicable if their standard proposal is accepted in its entirety, and contend that the group expects that there will be enough changes in the proposal to allow them to wiggle out of the zero-fee model at their convenience. Theoretically, at least, WWiSE's licensees could end up under a pay-for-play RAND agreement that was no different from the TGnSync IP licensing arrangement that Atheros is a party to.
I find both the legal and technical arguments very interesting -- even if some of them remind me a bit of the misleading legalistic nit-picking being used by some of the spin-meisters in the current Presidential campaign. As with any good standards debate, both sides make compelling arguments, ones that that are most often equal mixes of genuine technical idealism and artful sophistry designed to disguise a more self-serving agenda.
As my busy weekly schedule permits, I'll research these issues and do my best to verify or debunk the claims being made here. In the meanwhile, I've offered Atheros a chance to publish a second position paper here to discuss their position on this, and other aspects of the 802.11n standard. As always, the WWiSE group, and any other responsible party will also have a place to air their point of view.
Meanwhile, I'll keep trying to call things as I see them, admit when I get things wrong, and rely on the enthusiastic inputs from my readers and other friends to eventually get things right.
Comments? Questions? Eagle-eyed criticisms? I'll be waiting here every Monday morning, coffee cup in hand, to read your feedback at: lgoldberg@green-electronics.com