Scott Rafer: Wi-Max will join HomeRF in Intel’s dustbin of proprietary wireless standards by mid-2008.


DG Lewis

OK, I’m still asking: How do sub-$400 WiMAX-equivalent access points get the world to broadband internet for less than $20/month? How does this self-installing, self-provisioning, self-organizing mesh network actually connect to Level3, AT&T, MCI, and the other backbone providers? And, more significantly, how does it pay to connect to them?

Jesse Kopelman

No way real (IEEE) 802.11n is available before real 802.16e. The 802.16e standard will be done and ready for certification by the end of this year, not true for 802.11n.

Scott Rafer

Finishing the move from AO to an environment where the conversation is exposed to a focused audience:

From JCH ——–
[jch] | POSTED: 11.09.05 @14:00

The assumption is not layered as you would think, and certainly not a rural vs urban model here. No cogent alternatie speaks not to free vs user subscribed but rather how wireless broadband can be monetized in particular markets; clearly this is not uniformed as it might appear. Urban mesh networks are coming fast not because Craig McCaw can do it……[although he certainly is in urban markets], but because you can cheaply monetize the service faster than premium service RBOCs broadband and sets himself up for offering OFDM/WiMAX VoIP……like Sprint-Nextel soon will do. How do you compete with FTTH that SBC [soon AT&T], and Verizon are currently rolloing out in densely populated urban areas? If I move from NYC, LA or SF to Corpus Christi TX, Fort Myers FL….would I want broadband that matches the unique lifestyle features of the local area: you bet? [human/geographical factors]

McCaw’s Clearwire must not only offer just premium broadband but content……and that includes IPTV. So voice, video, data, and the voice is mobile as well as stationary. But please don’t knock Tony’s AO …….I’ll check out your WhyMAX? post and summarily rip it to shreds…..;-)

————- Rafer:———–

Aha, “monetize!” My point is that new carriers can’t monetize at a level that covers fixed and startup costs. This conversation got restarted after a year do to a column i just wrote called Wi-Fi is Not a Communications Service. That’s exactly it; ClearWire’s market spoiler is a loss leader, not a competitive service.

Maybe our specific problem is the usage of the term “cogent.” I’m entirely indifferent to single sign-on, carrier-provided security (as opposed to employer-provided security, etc.) I don’t think people care whether they are on different providers almost continuously as long as they feel safe, don’t spend a lot of time getting online, and get three 9’s (or maybe even two 9’s) of reliability. Beyond that standard the mass of the population won’t pay service premiums.

You’ve also surfaced a basic religious difference between us. All network carriers eventually fail at the content game — http://isen.com/stupid.html AT&T wanted to be the search and consumer auction site. The link above is why they failed. The mobile guys are not different than the copper LATA or fiber crowd. Their packet pushers who are having delusions of grandeur when they do anything else. I’m talking about a decades long process in the case of incumbent carriers, but once again — new carriers can’t break in that way. Therefore, the FTTH comment works in my favor doesn’t it? I’m not saying free Wi-Fi wins in all cases. I’m just saying it prevents new carriers from becoming profitable.

Scott Rafer

Hi Jeff, Thanks for saying so. I agree with you. Plus when 802.11n is made into a mesh (which is likely sooner than 802.16e ships in commercial products), it’s going to reinforce my points.

Scott Rafer

Hi DG, my and JCH’s responses (cross-posted from AO) are:

My dear chap, Craig McCaw hasn’t failed in wireless yet…………as oppossed to your analysis, but more to the point: you love to talk on the level of inferences of technology whilst ignoring the gist of the productive substance. Let me explain: you say “Will VOIP only dominate where voice services don’t yet exist?; that’s not the arguement. We’re not speaking metaphorically as referenced by my previous post……….only EV-DO and OFDM WiMAX/4G can undercut this. You love to take a time place snap-shot and call it reality whilst reality is being formed right front of your eyes. Clearwire is not buying exclusivity……..they are buying mobility…is that not evident? They are betting beyond a horizon that you haven’t referenced yet [or see for that matter]…VoIP on WiMAX using OFDM……….[study it: if you dare]

From Rafer:
You’re right. It’s only a metaphor. That’s why i specifically called it a “metaphor.”

You are layering your assumptions (“Unlicensed wireless will dominate where there are simply no cogent alternatives.”) onto my argument. The two don’t mix well. I’m completely focused on urban and relatively dense suburban environments, where the huge majority of the US population lives. And, where the US is 16th in the world (at best) in terms of broadband throughput and pricing. I appreciate that rural broadband is a important social, political, and technology issue, but it’s not one that I choose to work on.

McCaw hasn’t failed in terrestrial, mobile wireless yet. However, he has failed in broadband at least once. If McCaw can make ClearWire profitable purely on the hardcore mobile premium (faster than walking speeds only), then he might make it. Those prices will crash much later. However, I don’t believe he can form a business simply off that specific premium, making little or no money. Otherwise, he’s starting off with an Innovator’s Dilemma — cheaper, low-feature, good-enough solutions exist now and are being very quickly adopted. EVDO and WiMax have some hope slowing the ARPU decline for exisiting carriers (VZW and Sprint), but a new carrier won’t be able to pay off its startup costs.

You seem to assume my analysis is flawed because I prioritize economics and human factors over speeds and feeds. Feel free to continue. It makes competing with you easier.

Finally, if you wish to continue this, let’s do so in public. Being behind Tony’s security here at AO keeps out most of the people that would add value to this dialog. The best place is likely GigaOm [ http://gigaom.com/2005/11/08/whymax/#comments ] as it’s got the most traffic for this sort of thing: http://tinyurl.com/7vodu

—-and add’l from Rafer—-
Oh, sorry. I just realized that I mixed the comments of JCH and DG Lewis in my urban vs. rural remark. I apologize that my tone was a little over the top in that case. However, it seems like you guys largely agree, so I think the content of the response still works.

DG Lewis


A question I asked on the comments board at AlwaysOn, but per your suggestion, I’ll repeat it here.

How do all these user-installed and operated mesh networks connect to the rest of the world?

Let’s in fact assume that by next July I’ll be able to buy a mesh radio in Best Buy for a few hundred bucks (or I’m willing to spend $435 to buy a MeshBox from Locust). And let’s assume that a dozen other people in my neighborhood do likewise.

Oh, and by the way, let’s assume that I live in West Podunk, Wisconsin, where the telco doesn’t offer DSL service and the cable company doesn’t even run cable, let alone offer high-speed internet.

So now how do I get to Yahoo? Or gigaom.com? The nearest ISP POP is maybe 25 miles down the road. And they don’t have a MeshBox. And even if they did, they’re not going to serve all this traffic from my neighbors and me out of the goodness of their hearts.

Even if you focus on urban/dense suburban environments, how, exactly, does a user-owned and operated wireless mesh network provide connectivity to the internet, unless one or more of the users have an existing internet connection that they’re willing to share? And if they’re paying for that internet connection, how willing will they be to share it? Doesn’t it reduce to a Tragedy of the Commons?

Jeff Blaine

Scott, FWIW, my comment was the first in the list and unrelated to Om/you.

I bugged Om about his comment display author attribution being awkward :)

Jeff Hearon

I thought I’d have to do some real debunking here but Scott has debunked himself…..the apples and oranges comparison above clearly indicate this.

Ok boys and girls, once again:

The current 802.16 standard is IEEE Std 802.16-2004, approved in June 2004. It renders the previous (and 1st) version 802.16-2001 obsolete, along with its amendments 802.16a and 802.16c.

IEEE Std 802.16-2004 addresses only fixed systems. An amendment 802.16e is in the works which adds mobility components to the standard. This amendment is expected to be completed in mid 2005.

WiMAX’s Imminent Disruption
Forget the analysis look at the evidence

Why Ebay Acquired Skype: Wireless Broadband Wars Have Begun
Google enters WiFi arena

Scott Rafer

My point is that new carriers can’t cross into profitability where the user perception of their service is that it competes with the face of standards-based solutions on unlicensed bands.

On the hardware availability and cost issue:
1. Crack open one of those modems and you’ll find something rather similar to a low-end wireless access point. Are its differences from a low-end access point the hardware components or the software that locks you to McCaw’s backhaul? Being licensed gear, it’ll be higher power than unlicensed, but that’s not the point I’m seeking to make.
2. I’m guilty of believing some of the WiMax hype in spite of my criticisms. I thought the standards would be a lot further along than this: http://wimaxnetnews.com/archives/2005/10/understand_the.html My specific mistake was including a year in that statement. Here’s an attempt at a rewrite with the same logic but without my data-specific screwup that everyone is focusing on:

Old paragraph as found on: http://www.alwayson-network.com/comments.php?id=P4333_0_4_0_C
Once again, it’s just wireless Ethernet. There are no barriers to entry or economies of scale. Customer-deployed equipment will dominate, making WiMax-specific service providers unsustainable. I predict that WiMax access points will be sold for less than $400 on Amazon in 2005, and for less than $200 in 2006. Not only that, but they’ll be far easier to install than satellite TV dishes, and I bet they’ll reduce the total monthly fees for broadband Internet—including flat-rate global voice services—to less than $20 per month.

Suggested replacement paragraph, changes in italics:
Once again, it’s just wireless Ethernet. There are no barriers to entry or economies of scale. Customer-deployed equipment will dominate, making WiMax-specific service providers unsustainable. I predict that open-standards, backhaul-independent, unlicensed WiMax-equivalent access points will be sold for less than $400 at retail before ClearWire can get anywhere near a critical mass of subscribers (which is long before profitability). Not only that, but they’ll be far easier to install than satellite TV dishes, and I bet they’ll reduce the total monthly fees for broadband Internet —- including flat-rate global voice services —- to less than $20 per month.

The writing is a bit clunkier, but it’s the right conomic argument. What year it happens in is somewhat orthogonal to the argument, which I should have stated two years ago when I wrote it.

DG Lewis


What Clearwire is selling at Best Buy is not a WiMAX access point, it’s a modem. (It’s also not WiMAX, but that’s a quibble.) For it to do anything, Clearwire has to deploy NextNet Expedience base stations. They cost lots more than $400, and you can’t buy them at Best Buy.

Given that your point was that a WiMAX service provider model would be unsustainable because WiMAX access points would be so cheap that “customer-deployed equipment will dominate,” it’s a bit disingenuous for you to try to prove this point by citing the availability of a Broadband Wireless modem that requires subscription to a Broadband Wireless service provider.

DG Lewis

Personally, I think Scott’s predictive credibility is impaired by the article he references, which he wrote in June of 2004, in which he predicted that “WiMax access points will be sold for less than $400 on Amazon in 2005.”

Jeff Blaine

Off-topic, but for crying out loud, I just want official 802.11n products. I have painfully resisted buying any wireless gear for 3 years now. Surely the big players are tired of nerd-slapping each other about the spec details by now, right?

Comments are closed.