You're so wrong. These lights are governed by Part 18, ISM. They do not have be concerned about interference as long as they conform to the Part 18 rules, which are much broader than Part 15.
Part 18 governs industrial, scientific, and medical devices that don't communicate data, but rather emit radiation as a direct purpose of their utility: microwave offices, industrial sealers, etc.
I'm one of the plaintiffs in the suit. There is some risk for us and some inconvenience, but it's minor compared to the scale of the outcome if we win or if the "entertainment oligopoly" (as defined in the suit) backs down. I mean, I'm not a thief. I don't like being called a thief. And I'm willing to stand up to the bullies who are trying to take away a combination of constitutional and statutory rights and privileges. Copyright is a grant in the public interest; fair use allows a lot of leeway in individual private use of copyrighted materials.
This website pops up 64 popups every time you visit it. Which is likely to be once, after you realize your mistake, you won't come back.
Note that in this case, it's a typo-net system: if you type oogle or gogle instead of google, for instance. I type about 90 wpm, but I still wind up entering gogle, microsfot, etc. Even if each of us only makes one typo every two days in entering a popular Web site URLs, that's a whole lot of mistakes.
Verizon Wireless appears to actually care (in the most hard-nosed, financial sense of using customer service) about its customers. I've been using them since the Airtouch days, and have had mixed experiences with service and billing. However, a couple of times recently, I've had, well, extraordinary customer service people on the phone who practically begged to come to my home and apologize for minor problems.
I'm on a $120/month plan with tax (for home/office/roaming, 900 minutes, nationwide). Perhaps I'm escalated into a better support category. But still.
The other day, I was off to Canada for a few days, and I called Verizon to see if I could get a Canada roaming plan. Not only did they have one ($10/month) but they would pro-rate it just for when I needed it and automatically turn it off. Zounds.
Something went wrong, though, and it was applied for just three days instead of two weeks. I called up, they found the record, apologized profusely, added another note to their system, and said there was no way I'd be charged anything extra.
Amazing. Weird. It's not the customer's fault. What a strange idea.
Isbn.nu is my site; thanks for reffing it. Part of my goal in the site was to take a commercial source of book info (Baker and Taylor, in my case) and turn it into something that could be quickly used to find books.
The biggest problem with book information is that there's so damn much of it. The set is too large for individuals to contriubte enough. It grows rapidly and in a distributed fashion (at least 40,000 publishers in the US, and possibly more when you count every single-book publisher; 25,000 new books per year, or perhaps more, depending on how you count it). There are entrenched interests who actually spend millions per year collecting and correcting book info, too.
I'd be more interested in developing a resource that could be shared except that I know how many millions are spent.
The best bet is that publishers could be persuaded to form a non-profit group, the goal of which would be to collect and maintain information on books and disseminate it accurately to all the places that use book info. But that's unlikely, again, because of the cost.
(Formerly, I ran the catalog department at Amazon.com in 96-97, and have consulted for a couple of bookstores, including Powells and Half.com. The problem with book info is a Problem with a capital P.)
There's also SOHO Wireless. All three companies are offering variants on software-based hot spots in which you can take relatively cheap existing equipment and turn it into a commercial-grade public access hot spot without building back-office billing, support, network, etc. I wrote up an analysis of the three companies back on March 11th on my Wi-Fi blog at http://80211b.weblogger.com/2002/03/11/
So for $900 I get an unnetworked machine lacking FireWire but probably including USB 1.0 or 1.1, no Wi-Fi antenna or card, and no monitor. For $1,400, I get a 15" LCD, two FireWire ports, three USB 1.1 ports, Wi-Fi antenna built into the case (+$100 for internal card), 10/100 Mbps Ethernet, and the equvalent of a 1.5 GHz P4. I'm not seeing how the PC is cheaper for people who actually *do* things with their computers like move data around.
Actually, the good news is that Microsoft and Intel, Apple and Agere (Lucent spinoff), Intersil and Texas Instruments, Sprint PCS and VoiceStream, and others, all have a vested financial interest in keeping 2.4 GHz open for unlicensed use as they sell, resell, or charge for billions of dollars of equipment or service.
You're not incorrect, it's just that the chipmakers want to offload host processing to the chipsets instead of relying on host processing. This assures more symmetric performance across all machines and in network intensive situations in which you might be running 802.11a (54 Mbps raw), sucking host computational power may not be appropriate. For smaller devices that are Compact Flash based and use ARM processors or similar devices, host processing would be in appropriate and require a separate chipmaking development track.
The OEMs, likewise, don't want to pay and support driver development for host-based processing. They want to feed data into an onboard MAC that offloads it to the on-board encryption which sends it off to the PHY and radio.
You're absolutely right on all the other fronts, of course: AES is a piece of the puzzle and there's not a specific reason to not use host processing; it's a gestalt.
As for the rest, the industry hasn't yet taken that deep breath and said, we have to rethink the problem end-to-end.
Folks, Google is making scads of money selling search service to business. Go to Cisco and other sites, and you see "powered by Google." They make a lot of money off this service because, even as an outsourced service, they can save hundreds of thousands of dollars a year in staff, maintenance, and software licenses and development from in-house search engine deployment.
The Google in a box service is just an extension of their existing business service. So while it's a great thing, it's another tool. They're already nearly or entirely profitable, apparently, between ads and their business search service.
Thank you. I can't believe with all the brilliant Slashdot minds it took this long for someone to post the sensible analysis!
It's all about speed versus hardware at Google: they will always add hardware to keep the speed of queries consistent. If you can add less hardware (possibly 10% of what you'd need with a hard disk dependency) by relying on RAM, you save a bundle and keep performance at a reasonable rate.
Also, depending on how linear the curve is, you throw another 1Gb in 100 machines instead of buying another 1,000 machines.
I had a guy threaten me several months ago because I wouldn't forward a post he wrote to a mailing I moderate (on a software topic: Adobe GoLive) in part because it was belligerant. Eventually, I banned him. He then threatened that he'd "get me."
Well, this was pre-Sept. 11, but I thought, I don't know where he lives, I don't know how crazy he is. I had his name and his email address. I sent a note to the system administrator of the address noting that threatening email had been sent from that location. (Turned out to be his work address.) I also used Switchboard.com to get his phone number.
I called. I got what I thought was his mom (I assumed he was about 18 up to that point), but turned out to be his wife. I said, if I received any additional communication or anything happened to me or my systems, I would be reporting him to his local police and FBI. She said she's pass the message on.
A few days later I get email begging me to never get in touch with him again. I felt slightly bad: did I want him to lose his job? No. But I didn't want to worry about about a random crazy (who turned out to live about 1,000 miles away) who might hack my systems or my body up.
Practically every post on this matter misses the point. Boingo is combining aggregation of existing network points with software. Yes, software. Good old software.
There are a ton of useful things that you can't do via a Web browser even with unreliable Java. All of these useful things for Wi-Fi require protocol-level decisions and interaction with software.
Boingo's client has built-in VPN support to their public POP (i.e., bypass WEP, bypass lack of WEP, bypass security problems in local hot spots on their wired portion); it uses NDIS 5.1 to sniff local networks; etc.
Where Boingo is going to make the market work is by opening up the networks to multiple players, of which they are just one. They will be the premier player (and certainly are at the moment).
iPass is the only other player to have real software behind their aggregation. But it's not designed as strictly for the Wi-Fi customer and the various needs one has in that market.
This is correct. 802.11b channelizes into 14 overlapping 22 MHz bands from 2.400 to 2.485 GHz (in the U.S.; fewer frequencies are available in some other countries, but all in that range). But there's no reserved "g" bandwidth.
This is a great point. Until Thursday, I thought that the Balkanization of high-bit-rate 2.4 GHz 802.11b-like flavors would keep 802.11b the dominant standard for compatibility. But now, you're right: cheap "b" will become even cheaper, as companies hold out for "g" equipment. Why spend money on "b" today given that "g" will only cost incrementally more, if that.
Good point, but OFDM and PBCC are both supposed to be much more adaptive and responsive to frequency interference, both temporary and permanent. My understanding of OFDM is that it has some properties in common with DMT (Discrete Multitone Transform) which is part of the wavelet-based algorithm for encoding in higher-speed DSL system. DMT channelizes available bandwidth into smaller chunks, each of which has a QAM-style modulation like an individual modem. So each channel can be responsive to errors and interference, and entire channels can be shut down without radically altering the entire bandwidth profile. Further, you can step down the whole chunk (all channels) to a lower speed and still benefit from channelization within that lower speed.
"Gotta learn to read through the marketing-speak."
Or, you could try to not be a Slashdot troll and actually read the flipping information.
I spoke to Apple about the AOL support, and they enable people with the Mac AOL 5.0 client software to relay through the Base Station. The Base Station directly dials AOL's servers with the client software acting as the engine that drives it. This is huge for millions of Mac/AOL users, even though it seems irrelevant for the digerati.
I recently was blocked from sending email to a colleague because it contained a URL embedded in the email that had numbers in the URL (80211b.weblogger.com).
The bounced message pointed me to a Web page that was supposed to explain why it was bounced. It didn't. I emailed the sysadmin of the ISP who asked to see my email that I'd sent. I objected: why should I show him private email to a colleague with whom I'd previously corresponded?
The sysadmin relented, figured out the problem (he was bouncing email that had those 32-bit numeric URLs in them), and fixed it. But still...
I agree almost entirely with you, Monster: you've struck a perfectly reasonable balance. You're not arguing against limits, just against arbitrarily non-consensually non-contractually applied limits.
I might argue more broadly that spamming is a social problem, but the only way to solve it is through technological cooperation and user involvement in the decision making.
I subscribe to MAPS for my sendmail server because it's the best solution and I don't have real end users (we're a small group of folks sharing a server).
802.11a runs up to about 150 feet indoors *at full speed* where 802.11b can run more like 300 feet. (These are just random numbers, of course, because internal obstacles like plaster coated chicken wire stops transmissions.)
But 802.11a has a number of step down speeds: if it can't do 54 Mbps, it drops to the next, and so on. I believe it has 12 stepdowns to 802.11b's 4 (1, 2, 5.5, 11).
This means that where 802.11b might be able to run at 1 Mbps at a few hundred feet from an access point, 802.11a could still be running at 12 Mbps.
Further, when you get out into the open landscape and can do point-to-point, you can run miles and miles, just as with 802.11b. Or, with an access point mounted externally for a neighborhood or campus.
And 802.11a uses the 5 GHz band, which is uncrowded and reserved, unlike 2.4 GHz (Bluetooth, HomeRF, cordless phone, microwave oven interference).
Actually, I wrote the article, and I'm not a biased Mainer. I live in Seattle; I lived in Maine just two years, and make it back every year or so. There are hundreds of thousands of people in Midcoast, split out into communities of anywhere from a few hundred to several thousand across large tracts of land and islands reachable only by private boat, or sometimes by ferry.
Why is this PR? I don't get your point of view. The fact is that some percentage of users need, want, and can afford high-speed access, but more reasonably, it's about businesses and entrepreneurs who can't run their careers or their operations without affordable Internet bandwidth.
I can tell you about a lot of bookstores around the country, for instance, for whom the Internet means an extra 5 to 10 percent revenue per year through online sales via ABEBooks.com and other independent, inexpensive venues. This is basically their profit margin, so making an extra few percentage points can mean the difference between a viable business or not.
Ultimately, the combination of what MIS is offering is the best part, and repeatable in other rural areas: using wireless as a funding source to bring service to ever more remote areas, where they can offer local dial-up (non toll) via local exchanges to modems, or even DSL through local telcos as they are in Damariscotta.
You're talking about abstract 802.11; I'm talking about the common implementation of it prior to the "b" spec, which uses FH for speeds of up to 3 Mbps raw with orthogonal hopping patterns.
You can't simply get the SSID by asking around because the customers are few and the equipment is ISP-style stuff: the customers don't configure the client side or have direct access to it.
Even if you manage to grab the SSID, secure the hopping pattern by making a connection, you still have to do an authentication as a legitimate user.
This is a case of "read the original article, please." I didn't say Portland's cable modem service was broken: I said that in Midcoast, the cable modem service was generally seen as flaky. Adelphia took over a number of local operators, and the Mint service, which worked okay until then, started to stop working okay. One person I spoke to has about one full day of cable modem service out of 30, down from almost 100% uptime previously.
Fiber is only available in limited parts of the state in limited places. "Coast to coast" doesn't help when you can't tie into it where you are. In the midwest, across most major routes, you've got vast amounts of dark fiber, and quite a lot of it lit up. It's trivial for Omaha to get gigabits to either coast. Not so trivial for Camden or Bar Harbor.
Always enjoyable to see the media "discover" something that's been around for a long while:)
A. I'm not the media. I'm one guy who is obsessed by 802.11 and all its letters. O'Reilly Network is a developer editorial site, not the Washington Post.
B. I clearly state in the article that they've been running this service since 1997. I thought their particular story, especially with four years of solid experience and their rejection of 802.11b in favor 802.11, were all interesting points.
I didn't pretend, nor did the site, that we discovered these guys. They were happy enough tooling along without any publicity.
The 5 GHz spectrum has been opened up to newer WLAN flavors, mostly 802.11a at present. There's a lot of unused spectrum in several ranges that won't conflict with existing flavors. The 802.11a encoding algorithm, OFDM, is more forgiving of signal reflection and overlap, too. It runs at 54 Mbps, but has about a dozen step-down slower speeds, so you can run at all kinds of rates in overlapping cells.
Slashdot Mirror
You're so wrong. These lights are governed by Part 18, ISM. They do not have be concerned about interference as long as they conform to the Part 18 rules, which are much broader than Part 15.
Part 18 governs industrial, scientific, and medical devices that don't communicate data, but rather emit radiation as a direct purpose of their utility: microwave offices, industrial sealers, etc.
I'm one of the plaintiffs in the suit. There is some risk for us and some inconvenience, but it's minor compared to the scale of the outcome if we win or if the "entertainment oligopoly" (as defined in the suit) backs down. I mean, I'm not a thief. I don't like being called a thief. And I'm willing to stand up to the bullies who are trying to take away a combination of constitutional and statutory rights and privileges. Copyright is a grant in the public interest; fair use allows a lot of leeway in individual private use of copyrighted materials.
Note that in this case, it's a typo-net system: if you type oogle or gogle instead of google, for instance. I type about 90 wpm, but I still wind up entering gogle, microsfot, etc. Even if each of us only makes one typo every two days in entering a popular Web site URLs, that's a whole lot of mistakes.
Verizon Wireless appears to actually care (in the most hard-nosed, financial sense of using customer service) about its customers. I've been using them since the Airtouch days, and have had mixed experiences with service and billing. However, a couple of times recently, I've had, well, extraordinary customer service people on the phone who practically begged to come to my home and apologize for minor problems.
I'm on a $120/month plan with tax (for home/office/roaming, 900 minutes, nationwide). Perhaps I'm escalated into a better support category. But still.
The other day, I was off to Canada for a few days, and I called Verizon to see if I could get a Canada roaming plan. Not only did they have one ($10/month) but they would pro-rate it just for when I needed it and automatically turn it off. Zounds.
Something went wrong, though, and it was applied for just three days instead of two weeks. I called up, they found the record, apologized profusely, added another note to their system, and said there was no way I'd be charged anything extra.
Amazing. Weird. It's not the customer's fault. What a strange idea.
Isbn.nu is my site; thanks for reffing it. Part of my goal in the site was to take a commercial source of book info (Baker and Taylor, in my case) and turn it into something that could be quickly used to find books.
The biggest problem with book information is that there's so damn much of it. The set is too large for individuals to contriubte enough. It grows rapidly and in a distributed fashion (at least 40,000 publishers in the US, and possibly more when you count every single-book publisher; 25,000 new books per year, or perhaps more, depending on how you count it). There are entrenched interests who actually spend millions per year collecting and correcting book info, too.
I'd be more interested in developing a resource that could be shared except that I know how many millions are spent.
The best bet is that publishers could be persuaded to form a non-profit group, the goal of which would be to collect and maintain information on books and disseminate it accurately to all the places that use book info. But that's unlikely, again, because of the cost.
(Formerly, I ran the catalog department at Amazon.com in 96-97, and have consulted for a couple of bookstores, including Powells and Half.com. The problem with book info is a Problem with a capital P.)
There's also SOHO Wireless. All three companies are offering variants on software-based hot spots in which you can take relatively cheap existing equipment and turn it into a commercial-grade public access hot spot without building back-office billing, support, network, etc. I wrote up an analysis of the three companies back on March 11th on my Wi-Fi blog at http://80211b.weblogger.com/2002/03/11/
So for $900 I get an unnetworked machine lacking FireWire but probably including USB 1.0 or 1.1, no Wi-Fi antenna or card, and no monitor. For $1,400, I get a 15" LCD, two FireWire ports, three USB 1.1 ports, Wi-Fi antenna built into the case (+$100 for internal card), 10/100 Mbps Ethernet, and the equvalent of a 1.5 GHz P4. I'm not seeing how the PC is cheaper for people who actually *do* things with their computers like move data around.
An earlier, seemingly legitimate post from a Red Hat employee said this isn't the 7.3 beta, but a pre-7.3-beta release (7.2.92?)
Actually, the good news is that Microsoft and Intel, Apple and Agere (Lucent spinoff), Intersil and Texas Instruments, Sprint PCS and VoiceStream, and others, all have a vested financial interest in keeping 2.4 GHz open for unlicensed use as they sell, resell, or charge for billions of dollars of equipment or service.
You're not incorrect, it's just that the chipmakers want to offload host processing to the chipsets instead of relying on host processing. This assures more symmetric performance across all machines and in network intensive situations in which you might be running 802.11a (54 Mbps raw), sucking host computational power may not be appropriate. For smaller devices that are Compact Flash based and use ARM processors or similar devices, host processing would be in appropriate and require a separate chipmaking development track.
The OEMs, likewise, don't want to pay and support driver development for host-based processing. They want to feed data into an onboard MAC that offloads it to the on-board encryption which sends it off to the PHY and radio.
You're absolutely right on all the other fronts, of course: AES is a piece of the puzzle and there's not a specific reason to not use host processing; it's a gestalt.
As for the rest, the industry hasn't yet taken that deep breath and said, we have to rethink the problem end-to-end.
Folks, Google is making scads of money selling search service to business. Go to Cisco and other sites, and you see "powered by Google." They make a lot of money off this service because, even as an outsourced service, they can save hundreds of thousands of dollars a year in staff, maintenance, and software licenses and development from in-house search engine deployment.
The Google in a box service is just an extension of their existing business service. So while it's a great thing, it's another tool. They're already nearly or entirely profitable, apparently, between ads and their business search service.
Thank you. I can't believe with all the brilliant Slashdot minds it took this long for someone to post the sensible analysis!
It's all about speed versus hardware at Google: they will always add hardware to keep the speed of queries consistent. If you can add less hardware (possibly 10% of what you'd need with a hard disk dependency) by relying on RAM, you save a bundle and keep performance at a reasonable rate.
Also, depending on how linear the curve is, you throw another 1Gb in 100 machines instead of buying another 1,000 machines.
I had a guy threaten me several months ago because I wouldn't forward a post he wrote to a mailing I moderate (on a software topic: Adobe GoLive) in part because it was belligerant. Eventually, I banned him. He then threatened that he'd "get me."
Well, this was pre-Sept. 11, but I thought, I don't know where he lives, I don't know how crazy he is. I had his name and his email address. I sent a note to the system administrator of the address noting that threatening email had been sent from that location. (Turned out to be his work address.) I also used Switchboard.com to get his phone number.
I called. I got what I thought was his mom (I assumed he was about 18 up to that point), but turned out to be his wife. I said, if I received any additional communication or anything happened to me or my systems, I would be reporting him to his local police and FBI. She said she's pass the message on.
A few days later I get email begging me to never get in touch with him again. I felt slightly bad: did I want him to lose his job? No. But I didn't want to worry about about a random crazy (who turned out to live about 1,000 miles away) who might hack my systems or my body up.
Practically every post on this matter misses the point. Boingo is combining aggregation of existing network points with software. Yes, software. Good old software.
There are a ton of useful things that you can't do via a Web browser even with unreliable Java. All of these useful things for Wi-Fi require protocol-level decisions and interaction with software.
Boingo's client has built-in VPN support to their public POP (i.e., bypass WEP, bypass lack of WEP, bypass security problems in local hot spots on their wired portion); it uses NDIS 5.1 to sniff local networks; etc.
Where Boingo is going to make the market work is by opening up the networks to multiple players, of which they are just one. They will be the premier player (and certainly are at the moment).
iPass is the only other player to have real software behind their aggregation. But it's not designed as strictly for the Wi-Fi customer and the various needs one has in that market.
This is correct. 802.11b channelizes into 14 overlapping 22 MHz bands from 2.400 to 2.485 GHz (in the U.S.; fewer frequencies are available in some other countries, but all in that range). But there's no reserved "g" bandwidth.
This is a great point. Until Thursday, I thought that the Balkanization of high-bit-rate 2.4 GHz 802.11b-like flavors would keep 802.11b the dominant standard for compatibility. But now, you're right: cheap "b" will become even cheaper, as companies hold out for "g" equipment. Why spend money on "b" today given that "g" will only cost incrementally more, if that.
Good point, but OFDM and PBCC are both supposed to be much more adaptive and responsive to frequency interference, both temporary and permanent. My understanding of OFDM is that it has some properties in common with DMT (Discrete Multitone Transform) which is part of the wavelet-based algorithm for encoding in higher-speed DSL system. DMT channelizes available bandwidth into smaller chunks, each of which has a QAM-style modulation like an individual modem. So each channel can be responsive to errors and interference, and entire channels can be shut down without radically altering the entire bandwidth profile. Further, you can step down the whole chunk (all channels) to a lower speed and still benefit from channelization within that lower speed.
"Gotta learn to read through the marketing-speak."
Or, you could try to not be a Slashdot troll and actually read the flipping information.
I spoke to Apple about the AOL support, and they enable people with the Mac AOL 5.0 client software to relay through the Base Station. The Base Station directly dials AOL's servers with the client software acting as the engine that drives it. This is huge for millions of Mac/AOL users, even though it seems irrelevant for the digerati.
I recently was blocked from sending email to a colleague because it contained a URL embedded in the email that had numbers in the URL (80211b.weblogger.com).
The bounced message pointed me to a Web page that was supposed to explain why it was bounced. It didn't. I emailed the sysadmin of the ISP who asked to see my email that I'd sent. I objected: why should I show him private email to a colleague with whom I'd previously corresponded?
The sysadmin relented, figured out the problem (he was bouncing email that had those 32-bit numeric URLs in them), and fixed it. But still...
I agree almost entirely with you, Monster: you've struck a perfectly reasonable balance. You're not arguing against limits, just against arbitrarily non-consensually non-contractually applied limits.
I might argue more broadly that spamming is a social problem, but the only way to solve it is through technological cooperation and user involvement in the decision making.
I subscribe to MAPS for my sendmail server because it's the best solution and I don't have real end users (we're a small group of folks sharing a server).
Read the press release and post your comment, eh?
802.11a runs up to about 150 feet indoors *at full speed* where 802.11b can run more like 300 feet. (These are just random numbers, of course, because internal obstacles like plaster coated chicken wire stops transmissions.)
But 802.11a has a number of step down speeds: if it can't do 54 Mbps, it drops to the next, and so on. I believe it has 12 stepdowns to 802.11b's 4 (1, 2, 5.5, 11).
This means that where 802.11b might be able to run at 1 Mbps at a few hundred feet from an access point, 802.11a could still be running at 12 Mbps.
Further, when you get out into the open landscape and can do point-to-point, you can run miles and miles, just as with 802.11b. Or, with an access point mounted externally for a neighborhood or campus.
And 802.11a uses the 5 GHz band, which is uncrowded and reserved, unlike 2.4 GHz (Bluetooth, HomeRF, cordless phone, microwave oven interference).
Actually, I wrote the article, and I'm not a biased Mainer. I live in Seattle; I lived in Maine just two years, and make it back every year or so. There are hundreds of thousands of people in Midcoast, split out into communities of anywhere from a few hundred to several thousand across large tracts of land and islands reachable only by private boat, or sometimes by ferry.
Why is this PR? I don't get your point of view. The fact is that some percentage of users need, want, and can afford high-speed access, but more reasonably, it's about businesses and entrepreneurs who can't run their careers or their operations without affordable Internet bandwidth.
I can tell you about a lot of bookstores around the country, for instance, for whom the Internet means an extra 5 to 10 percent revenue per year through online sales via ABEBooks.com and other independent, inexpensive venues. This is basically their profit margin, so making an extra few percentage points can mean the difference between a viable business or not.
Ultimately, the combination of what MIS is offering is the best part, and repeatable in other rural areas: using wireless as a funding source to bring service to ever more remote areas, where they can offer local dial-up (non toll) via local exchanges to modems, or even DSL through local telcos as they are in Damariscotta.
How is this PR?
This is frustrating, because you're speculating.
You're talking about abstract 802.11; I'm talking about the common implementation of it prior to the "b" spec, which uses FH for speeds of up to 3 Mbps raw with orthogonal hopping patterns.
You can't simply get the SSID by asking around because the customers are few and the equipment is ISP-style stuff: the customers don't configure the client side or have direct access to it.
Even if you manage to grab the SSID, secure the hopping pattern by making a connection, you still have to do an authentication as a legitimate user.
This is a case of "read the original article, please." I didn't say Portland's cable modem service was broken: I said that in Midcoast, the cable modem service was generally seen as flaky. Adelphia took over a number of local operators, and the Mint service, which worked okay until then, started to stop working okay. One person I spoke to has about one full day of cable modem service out of 30, down from almost 100% uptime previously.
Fiber is only available in limited parts of the state in limited places. "Coast to coast" doesn't help when you can't tie into it where you are. In the midwest, across most major routes, you've got vast amounts of dark fiber, and quite a lot of it lit up. It's trivial for Omaha to get gigabits to either coast. Not so trivial for Camden or Bar Harbor.
Always enjoyable to see the media "discover" something that's been around for a long while :)
A. I'm not the media. I'm one guy who is obsessed by 802.11 and all its letters. O'Reilly Network is a developer editorial site, not the Washington Post.
B. I clearly state in the article that they've been running this service since 1997. I thought their particular story, especially with four years of solid experience and their rejection of 802.11b in favor 802.11, were all interesting points.
I didn't pretend, nor did the site, that we discovered these guys. They were happy enough tooling along without any publicity.
The 5 GHz spectrum has been opened up to newer WLAN flavors, mostly 802.11a at present. There's a lot of unused spectrum in several ranges that won't conflict with existing flavors. The 802.11a encoding algorithm, OFDM, is more forgiving of signal reflection and overlap, too. It runs at 54 Mbps, but has about a dozen step-down slower speeds, so you can run at all kinds of rates in overlapping cells.