You're exactly right, although a P2P network would only be part of it: someone without access to the client software should still be able to download the book. Or I can encourage folks using the P2P network to also host. That's started to happen. The ancient Info-Mac archive agreed via a colleague to host the file, so it's already been mirrored onto 20 or 30 sites, and I'm using a script to round-robin select one of those locations for the current download page.
I'll be making a tiny tiny amount of money writing about this for O'Reilly Networks, but there's a huge gap between what you can get by writing and the amount of the bandwidth bill. I have to hope Level 3 has humanity.
Interesting to see this story, because I just had a disaster in giving away the electronic edition of Real World Adobe GoLive 6. Peachpit Press published the book in March 2002, and we had the rights to release it electronically, for fee or free, and with the sales of the title low, we decided to give it away.
Unfortunately, I hosted the book on a server run by a friend at a Level 3 co-location, which charges by the 9th busiest hour. In 36 hours, we had 10,000 downloads of an average of 20 Mb each. Right. So we hit potentially a $15,000 bill for the ninth busiest hour being 16 Mbps (the first 1 Mbps was included in his monthly bill).
So I'm screwed here, of course, and trying to raise a dollar or two from folks who downloaded the book and found it useful. We don't know the final bill, and we don't know whether Level 3 will negotiate. This is more like a natural disaster than a business decision.
If I'd been smart, of course, I would have distributed the download to many sites with no bandwidth fees or limited numbers of simultaneous users. I just thought we'd get a few hundred downloads. Not 10,000.
This is a great question, and the Broadcom people are listeners. I'll be talking to some of them in a few days in my role as a gadfly about things wireless. I'll report back (on my blog or to this thread) about their take on Linux.
My understanding is that it can be awfully tricky to build Linux support for network devices that don't release any sort of API unless you have a dedicated team willing to reverse engineer calls, decompile software, etc. I don't know if there will be a critical mass to make that happen.
Since the first company that offers Linux support in any way will obviously be the one that gets the sales from the hundreds of thousands of Linux users who want G added to their laptops and handhelds, there's a clear financial motivation, no?
Interestingly, nobody in the article or here mentions the fact that there's a very active group at the IETF that's working on securing all kinds of authentication messaging systems, including EAP (the method for 802.1x wired/wireless authentication). EAP is the focus, but the papers presented at the various conferences cover many authentication methods and methods of securing them.
Protocols like EAP (Extensible Authentication Protocol) are intended to provide a generic mechanism over any transport system to handle legacy and modern handshaking and exchange to authenticate a user in a system.
In 802.1x/EAP, included as of the 802.11i wireless security update, 802.1x defines the roles of a client, access point authentication passthrough, and an authenticator. 802.1x restricts access to the network until the access point using EAP has been told by the authentication system that the client is okay to be on the network. It hands off a key, which eliminates spoofing, as even if you spoof the MAC address, you don't have the key. The key can be swapped frequently, like every 10,000 packets.
The problem with 802.1x/EAP is the same as with the SIM/GSM authentication system as described here. The authentication is sent in the clear! So you have three flavors of tunneled, SSL-like EAP: EAP-TLS (requires a pre-installed certificate on the client), EAP-TTLS (Meetinghouse, Funk support, tunnels EAP inside a tunnel), and PEAP (Microsoft, Cisco, same tunneling but ignores legacy protocols supported within EAP-TTLS).
When you compare PC and Mac laptops, you have to figure in from $100 to $350 in extra costs on the PC side to obtain the same amount of default battery life as the included iBook and PowerBook batteries.
The PC laptops will cycle way way down (so much for performance comparisons with their actually faster processors -- when plugged in) and shut off various parts to manage battery life while a Mac laptop, even without engaging in all the tricks, ekes out substantially better times.
Just ask Walt Mossberg of the Wall Street Journal and other PC-oriented journalists who review Macs, but typically don't use them. They get a Mac as a review unit to test and they rave about the battery, the screen, and the built-in Ethernet, FireWire, USB, VGA, DVI (some models), AirPort/Wi-Fi, etc.
It's really hard to find everything that you get out of the box in a Mac laptop in a single PC without an additional card or set of modules or an awkward port replicator on those smaller subcompacts.
Geez, I sound like an advertisement, but this is from real world experience. I used a Sony Vaio as my primary machine for well over a year when traveling and battery management was unbelievable even with a tiny screen and extra ports turned off and the LCD turned down -- and new batteries, not used ones.
I switched to a white dual-USB iBook when it came out, and suddenly I just don't worry about battery life at all any more.
$80 million spent, not lost or in debt
on
Salon Asks for Help
·
· Score: 4, Insightful
Let's be clear: some of the money they've raised over the year is simply gone, exchanged for stockholder equity: that's investment capital, not lost money. They spent it, the stockholders have stock, the money isn't owed to anyone.
A very small amount of money in relative terms is actually owed, under a few million I believe, but their operating costs exceed their income and they don't have any sources of stock -> money exchanges.
It's still ridiculous, of course, to have spent that much, but it's just pissed away not owed.
You're misreading their financials. If you read their SEC filings which rely on GAAP (generally accepted accounting principles) it's $4.2B and $400M.
If you read the link you pointed us to, shows $2.6B in cash/cash equivalents and $1.9B in short-term investments. I apologize for conflating short-term investments and cash/cash equivalents. Short-term investments are categorized that way as there is little risk or cost in liquidating those positions.
If you want to avoid GAAP and look at overall short-term money and expenses it's about $6B plus and $2B minus.
How are you getting your numbers from that link? By ignoring everything you don't like?
Apple has never had 9 billion that I know of. They've had about four billion for a few years. (The New York Times once misreported that they had 11 billion, and I ragged them via email until they printed a correction. I had to point out the time mark in the QuickTime broadcast of the financial conference call to prove it to them...)
Thanks for the great laugh. Apple has $4.2 billion (4,200,000,000 = 10 digits) in cash and cash instruments. They have about $400M in obligations. They also own a big piece of Akamai, still have ARM Holding stock, and have small holdings in many other firms which may still result in something more than their minimal early investments.
If you take the last few recession years, Apple has, overall not lost money, even though they've had small profits some quarters and small losses others. Even during the time they were investing heavily in research and development.
So how long does it takes a company with four billion dollars and no losses to go out of business? My calculator can't figure it out, but I'm off to use iMovie and Safari and iSync and AppleWorks and figure it out.
The approach for standardization is there, so we just have to hope that the companies come together.
I'm guessing we'll see an incredible fast track in July (if ratified then) or the next meeting with the Wi-Fi Alliance producing a ratified-802.11g test suite within months instead of longer.
Originally, the Wi-Fi Alliance was saying probably 2004 for a way to stamp Wi-Fi interoperability on top of 802.11g, but it seems ike it has to happen sooner.
The whole problem with shipping several companies' ideas of draft standards is that there's no central certification or testing, as there is with Wi-Fi. Several articles have said that Wi-Fi testing involves a plugfest. Well, there are plugfests, in which lots of manufacturers try interop with lots of devices, but there's also the Wi-Fi certification process with involves lab testing according to a long list of standards.
When 802.11g is finalized in a last draft soon, then is the time to buy 802.11g gear. I'm testing Linksys and Apple gear now, and although it's fine, there's no great motivation to hop on board until it all works correctly all the time.
InfoWorld reported this week on problems with speed, WEP compatibility, and cross-manufacturer compatibility. These will be fixed.
Hear, hear! If you read the New York Times Circuits story from two weeks ago about this, you find that the Everest part is just one aspect. What they're trying to do is bring the world to an area cut off by Marxist rebels that doesn't want to be cut off. Another way to avoid another Khmer Rouge killing fields situation is to have The Whole World Watching.
Wi-Fi is specifically a certified version of 802.11a or 802.11b or both. Not a generic term. It's confusing when used generically.
What's being discussed here is spread-spectrum frequency hopping or direct sequence -- probably FHSS not DSSS. It should be called by its right name as this is Slashdot, after all!
With 6 watts of output power, you could send 802.11b quite far as well using off-the-shelf equipment. This achievement is only remarkable because of the components involved (balloons, etc.), not because of the distance.
Because of the curvature of the earth, you have maximum distances without building huge towers that are only in the tens of miles, even with curving of the 2.4 GHz waves around the globe. The small wavelength means you don't get an enormous bending effect.
If you beam from the earth straight up, you have longer possible distances.
It's beyond just the middle, unallocated 5 GHz range: the military is criticizing the lower and upper band U-NII allocations, too.
The headline on this article is wrong: it's 802.11a, not 802.11b, that operates in 5 GHz. I confirmed with the NY Times journalist that 5 GHz was the only band at question here.
AltaVista almost freaked my wife out when we met
on
Googling For Dates?
·
· Score: 5, Interesting
In the dim recesses of Internet memory, AltaVista was king. I was working for Amazon.com, and a mutual friend suggested that the woman who is now my wife give me a ring to talk about working for a dotcom.
We met, hit it off, started dating, and five years later (this last Labor Day), got married.
Some dates after we met, she told me that she looked me up on AltaVista after she'd met me, and found 40,000 matches. (I was moderating the Internet Marketing Discussion List, www.i-m.com, and my name appeared on every post in the archives, which themselves appeared to be at many different domains.)
She said, if I'd looked you up beforehand I never would have called you. She would have been intimidated.
This is my experience as well. I'm in Seattle, so Speakeasy is local, but it's not why they've been so great. I was using Qwest at home and another local ISP for DSL at the office, and Qwest sucked (five days of downtime once, vast phone waits, untrained tech support) and the other ISP was charging by the gig for traffic!
I switched to Speakeasy and while it hasn't been entirely error free, they deal with their mistakes or the telco mistakes. When I moved my office a few months ago, I called and said, can we have a new circuit ready to go at the new location and then we switch the ATM circuit from my current office to my new office keeping all the IPs and so forth?
No problem they said. But we're moving on July 5 (a Friday after a Thursday holiday). No problem, they said. I called them, it took a few minutes, the IPs were moved. I shut boxes down at old office, drove them across town, powered them up, and I was live.
That's what the extra dough is for.
Because of the noncompetitive structure under which they have to get service, they're charging a realistic amount. When the next wave of technology comes, however, we should be seeing huge speed bumps in DSL without (in any reasonable world) the same kind of huge price increases.
Up north of Seattle in the tiny town of Snohomish, you can apparently get 8 Mbps down and 1 Mbps up for less than $100/month. Why? They brought fiber (or lit fiber) into the local switch, and the part of the telco setting prices up there thought this would work out. That puts the lie to the real underlying costs...
What are you, a phone company shill? This is packet data, to quote Dilbert. 11 Mbps is the raw data rate. 7 Mbps is an ideal real throughput after overhead. 4 Mbps is a realistic speed for a moderately used network using consumer gear.
In any case, though, users are bursty. If everyone wants to download video at the same time, bzzt, failure. Email, Web pages, etc., are not timebound for each packet and aren't simultaneous activities.
I've been using 10 Kbps GSM on the road the last couple of weeks, and for email, it rocks (versus expensive hotel phone calls and other options). It's all about what you need to do when and how fast.
Oddly, I wrote the first draft of what became that patent filing at Jeff's behest when I worked at Amazon.com from 1996 to 1997. Can't even recall why I was asked, but following the form of patent applications, I wrote out the ideas. Since I was an employee and not the inventor, I signed over all my rights. I was also under trade secrets restrictions because of my employment agreement.
I always wondered what happened to that patent, and lo and behold, here it is! It's certainly nostagia for me, but I was a pretty naive guy about patents in those days, and I wrote the draft not as a lawyer, but just as a technical guy who understand the mechanisms.
I have a very different opinion of things today, although it was clear at the time I wrote it that what Jeff had come up with was, in fact, unique, original, and significant under the way the law is still interpreted.
Most of the public-space community folks I've talked to around the country are not seeing anything like this kind of behavior, partly because community networks often (but not always) have enough people using them and aware of what's going on that it deters spammers and crackers. They're too exposed. They'd rather find unsecured mail servers than unsecured networks. They don't have a problem getting wireline access -- wireless isn't a big boon.
I mean, a cable modem in your house and some anonymous accounts work much better than freezing your tush in a park.
That said, there are a few simple things that could aid in making anyone trying to misuse a wireless connection, such as requiring a captive portal page a la nocat.net -- it scares people off. You can also use bandwidth restriction to make people don't abuse a segment. Sputnik.com has a free gateway software package that offers that option.
It's not about distance: it's about coverage. If you run point to point over a short distance, the signal becomes highly focused. These guys are not talking about boosting distances, but rather about increasing coverage without increasing ERP inside buildings, as one example.
If you had a 10,000 square foot facility and could put several access points (let's say 3 to 6) in a single server room with one of these antennas, you'd save a small fortune over, say, 20 to 30 access points all distributed across the actual area.
This makes perfect sense. If you had 20 clients on an AP with full bandwidth saturation (bad network designer, bad!), and 15 of them had good IVs as did the AP, and 5 had bad, then the amount of traffic it takes is now proportionally higher based on how many packets have good IVs attached. if I understood your explanation correctly.
Does anyone know which of the newer firmware uses improved IVs? As I understand it, you can use an AP with better IVs without having to have a wireless adapter with any changes in firmware. A few companies, mostly in the enterprise space, have made announcements about their changes, but I haven't seen anything that isn't proprietary between client/AP from a single vendor.
No, the WET11 can handle up to 30 devices and bridge them to any access point. The WAP11, when turned into client mode, replicates a single MAC address as a single-device bridge. If you've figured out how to bridge multiple devices with a WAP11 to any access point, call the newspapers!
You know a lot of the terminology, but I'm having trouble believing you read the article or looked at what the product does.
The WET11 can be used for a single device (hence the ad hoc mode option) or to bridge up to 30 devices to an AP wirelessly.
If you've got a bunch of wired devices plugged into a switch and want to connect to another wired network somewhere else that already has an AP attached, you can use this to bridge those devices onto the other network.
Very simple idea, very simple product. Most of what you're complaining isn't something the article does.
But where are you sniffing? You can't associate with the WET11 and sit actively on the network. You could sniff traffic going by, but the WET11 is likely to be used in short-range point-to-point installations with some sort of antenna unless it's entirely within someone's home.
I don't buy "WEP is crackable in a few days." That's a canard. It's absolutely crackable with sufficient data samples. From my reading and talking with security folks, home networks don't generate the amount of traffic necessary over short periods of time, like days, and corporate IT managers should be boiled alive if they're letting non-encrypted data pass over wireless links.
Slashdot Mirror
You're exactly right, although a P2P network would only be part of it: someone without access to the client software should still be able to download the book. Or I can encourage folks using the P2P network to also host. That's started to happen. The ancient Info-Mac archive agreed via a colleague to host the file, so it's already been mirrored onto 20 or 30 sites, and I'm using a script to round-robin select one of those locations for the current download page.
I'll be making a tiny tiny amount of money writing about this for O'Reilly Networks, but there's a huge gap between what you can get by writing and the amount of the bandwidth bill. I have to hope Level 3 has humanity.
Interesting to see this story, because I just had a disaster in giving away the electronic edition of Real World Adobe GoLive 6. Peachpit Press published the book in March 2002, and we had the rights to release it electronically, for fee or free, and with the sales of the title low, we decided to give it away.
Unfortunately, I hosted the book on a server run by a friend at a Level 3 co-location, which charges by the 9th busiest hour. In 36 hours, we had 10,000 downloads of an average of 20 Mb each. Right. So we hit potentially a $15,000 bill for the ninth busiest hour being 16 Mbps (the first 1 Mbps was included in his monthly bill).
So I'm screwed here, of course, and trying to raise a dollar or two from folks who downloaded the book and found it useful. We don't know the final bill, and we don't know whether Level 3 will negotiate. This is more like a natural disaster than a business decision.
If I'd been smart, of course, I would have distributed the download to many sites with no bandwidth fees or limited numbers of simultaneous users. I just thought we'd get a few hundred downloads. Not 10,000.
This is a great question, and the Broadcom people are listeners. I'll be talking to some of them in a few days in my role as a gadfly about things wireless. I'll report back (on my blog or to this thread) about their take on Linux.
My understanding is that it can be awfully tricky to build Linux support for network devices that don't release any sort of API unless you have a dedicated team willing to reverse engineer calls, decompile software, etc. I don't know if there will be a critical mass to make that happen.
Since the first company that offers Linux support in any way will obviously be the one that gets the sales from the hundreds of thousands of Linux users who want G added to their laptops and handhelds, there's a clear financial motivation, no?
Interestingly, nobody in the article or here mentions the fact that there's a very active group at the IETF that's working on securing all kinds of authentication messaging systems, including EAP (the method for 802.1x wired/wireless authentication). EAP is the focus, but the papers presented at the various conferences cover many authentication methods and methods of securing them.
Protocols like EAP (Extensible Authentication Protocol) are intended to provide a generic mechanism over any transport system to handle legacy and modern handshaking and exchange to authenticate a user in a system.
In 802.1x/EAP, included as of the 802.11i wireless security update, 802.1x defines the roles of a client, access point authentication passthrough, and an authenticator. 802.1x restricts access to the network until the access point using EAP has been told by the authentication system that the client is okay to be on the network. It hands off a key, which eliminates spoofing, as even if you spoof the MAC address, you don't have the key. The key can be swapped frequently, like every 10,000 packets.
The problem with 802.1x/EAP is the same as with the SIM/GSM authentication system as described here. The authentication is sent in the clear! So you have three flavors of tunneled, SSL-like EAP: EAP-TLS (requires a pre-installed certificate on the client), EAP-TTLS (Meetinghouse, Funk support, tunnels EAP inside a tunnel), and PEAP (Microsoft, Cisco, same tunneling but ignores legacy protocols supported within EAP-TTLS).
When you compare PC and Mac laptops, you have to figure in from $100 to $350 in extra costs on the PC side to obtain the same amount of default battery life as the included iBook and PowerBook batteries.
The PC laptops will cycle way way down (so much for performance comparisons with their actually faster processors -- when plugged in) and shut off various parts to manage battery life while a Mac laptop, even without engaging in all the tricks, ekes out substantially better times.
Just ask Walt Mossberg of the Wall Street Journal and other PC-oriented journalists who review Macs, but typically don't use them. They get a Mac as a review unit to test and they rave about the battery, the screen, and the built-in Ethernet, FireWire, USB, VGA, DVI (some models), AirPort/Wi-Fi, etc.
It's really hard to find everything that you get out of the box in a Mac laptop in a single PC without an additional card or set of modules or an awkward port replicator on those smaller subcompacts.
Geez, I sound like an advertisement, but this is from real world experience. I used a Sony Vaio as my primary machine for well over a year when traveling and battery management was unbelievable even with a tiny screen and extra ports turned off and the LCD turned down -- and new batteries, not used ones.
I switched to a white dual-USB iBook when it came out, and suddenly I just don't worry about battery life at all any more.
Let's be clear: some of the money they've raised over the year is simply gone, exchanged for stockholder equity: that's investment capital, not lost money. They spent it, the stockholders have stock, the money isn't owed to anyone.
A very small amount of money in relative terms is actually owed, under a few million I believe, but their operating costs exceed their income and they don't have any sources of stock -> money exchanges.
It's still ridiculous, of course, to have spent that much, but it's just pissed away not owed.
You're misreading their financials. If you read their SEC filings which rely on GAAP (generally accepted accounting principles) it's $4.2B and $400M.
If you read the link you pointed us to, shows $2.6B in cash/cash equivalents and $1.9B in short-term investments. I apologize for conflating short-term investments and cash/cash equivalents. Short-term investments are categorized that way as there is little risk or cost in liquidating those positions.
If you want to avoid GAAP and look at overall short-term money and expenses it's about $6B plus and $2B minus.
How are you getting your numbers from that link? By ignoring everything you don't like?
Apple has never had 9 billion that I know of. They've had about four billion for a few years. (The New York Times once misreported that they had 11 billion, and I ragged them via email until they printed a correction. I had to point out the time mark in the QuickTime broadcast of the financial conference call to prove it to them...)
Thanks for the great laugh. Apple has $4.2 billion (4,200,000,000 = 10 digits) in cash and cash instruments. They have about $400M in obligations. They also own a big piece of Akamai, still have ARM Holding stock, and have small holdings in many other firms which may still result in something more than their minimal early investments.
If you take the last few recession years, Apple has, overall not lost money, even though they've had small profits some quarters and small losses others. Even during the time they were investing heavily in research and development.
So how long does it takes a company with four billion dollars and no losses to go out of business? My calculator can't figure it out, but I'm off to use iMovie and Safari and iSync and AppleWorks and figure it out.
The approach for standardization is there, so we just have to hope that the companies come together.
I'm guessing we'll see an incredible fast track in July (if ratified then) or the next meeting with the Wi-Fi Alliance producing a ratified-802.11g test suite within months instead of longer.
Originally, the Wi-Fi Alliance was saying probably 2004 for a way to stamp Wi-Fi interoperability on top of 802.11g, but it seems ike it has to happen sooner.
The whole problem with shipping several companies' ideas of draft standards is that there's no central certification or testing, as there is with Wi-Fi. Several articles have said that Wi-Fi testing involves a plugfest. Well, there are plugfests, in which lots of manufacturers try interop with lots of devices, but there's also the Wi-Fi certification process with involves lab testing according to a long list of standards.
When 802.11g is finalized in a last draft soon, then is the time to buy 802.11g gear. I'm testing Linksys and Apple gear now, and although it's fine, there's no great motivation to hop on board until it all works correctly all the time.
InfoWorld reported this week on problems with speed, WEP compatibility, and cross-manufacturer compatibility. These will be fixed.
Draft, draft, draft!
Hear, hear! If you read the New York Times Circuits story from two weeks ago about this, you find that the Everest part is just one aspect. What they're trying to do is bring the world to an area cut off by Marxist rebels that doesn't want to be cut off. Another way to avoid another Khmer Rouge killing fields situation is to have The Whole World Watching.
Wi-Fi is specifically a certified version of 802.11a or 802.11b or both. Not a generic term. It's confusing when used generically.
What's being discussed here is spread-spectrum frequency hopping or direct sequence -- probably FHSS not DSSS. It should be called by its right name as this is Slashdot, after all!
With 6 watts of output power, you could send 802.11b quite far as well using off-the-shelf equipment. This achievement is only remarkable because of the components involved (balloons, etc.), not because of the distance.
Because of the curvature of the earth, you have maximum distances without building huge towers that are only in the tens of miles, even with curving of the 2.4 GHz waves around the globe. The small wavelength means you don't get an enormous bending effect.
If you beam from the earth straight up, you have longer possible distances.
It's beyond just the middle, unallocated 5 GHz range: the military is criticizing the lower and upper band U-NII allocations, too.
The headline on this article is wrong: it's 802.11a, not 802.11b, that operates in 5 GHz. I confirmed with the NY Times journalist that 5 GHz was the only band at question here.
In the dim recesses of Internet memory, AltaVista was king. I was working for Amazon.com, and a mutual friend suggested that the woman who is now my wife give me a ring to talk about working for a dotcom.
We met, hit it off, started dating, and five years later (this last Labor Day), got married.
Some dates after we met, she told me that she looked me up on AltaVista after she'd met me, and found 40,000 matches. (I was moderating the Internet Marketing Discussion List, www.i-m.com, and my name appeared on every post in the archives, which themselves appeared to be at many different domains.)
She said, if I'd looked you up beforehand I never would have called you. She would have been intimidated.
Thank goodness for a little lack of knowledge.
This is my experience as well. I'm in Seattle, so Speakeasy is local, but it's not why they've been so great. I was using Qwest at home and another local ISP for DSL at the office, and Qwest sucked (five days of downtime once, vast phone waits, untrained tech support) and the other ISP was charging by the gig for traffic!
I switched to Speakeasy and while it hasn't been entirely error free, they deal with their mistakes or the telco mistakes. When I moved my office a few months ago, I called and said, can we have a new circuit ready to go at the new location and then we switch the ATM circuit from my current office to my new office keeping all the IPs and so forth?
No problem they said. But we're moving on July 5 (a Friday after a Thursday holiday). No problem, they said. I called them, it took a few minutes, the IPs were moved. I shut boxes down at old office, drove them across town, powered them up, and I was live.
That's what the extra dough is for.
Because of the noncompetitive structure under which they have to get service, they're charging a realistic amount. When the next wave of technology comes, however, we should be seeing huge speed bumps in DSL without (in any reasonable world) the same kind of huge price increases.
Up north of Seattle in the tiny town of Snohomish, you can apparently get 8 Mbps down and 1 Mbps up for less than $100/month. Why? They brought fiber (or lit fiber) into the local switch, and the part of the telco setting prices up there thought this would work out. That puts the lie to the real underlying costs...
What are you, a phone company shill? This is packet data, to quote Dilbert. 11 Mbps is the raw data rate. 7 Mbps is an ideal real throughput after overhead. 4 Mbps is a realistic speed for a moderately used network using consumer gear.
In any case, though, users are bursty. If everyone wants to download video at the same time, bzzt, failure. Email, Web pages, etc., are not timebound for each packet and aren't simultaneous activities.
I've been using 10 Kbps GSM on the road the last couple of weeks, and for email, it rocks (versus expensive hotel phone calls and other options). It's all about what you need to do when and how fast.
Oddly, I wrote the first draft of what became that patent filing at Jeff's behest when I worked at Amazon.com from 1996 to 1997. Can't even recall why I was asked, but following the form of patent applications, I wrote out the ideas. Since I was an employee and not the inventor, I signed over all my rights. I was also under trade secrets restrictions because of my employment agreement.
I always wondered what happened to that patent, and lo and behold, here it is! It's certainly nostagia for me, but I was a pretty naive guy about patents in those days, and I wrote the draft not as a lawyer, but just as a technical guy who understand the mechanisms.
I have a very different opinion of things today, although it was clear at the time I wrote it that what Jeff had come up with was, in fact, unique, original, and significant under the way the law is still interpreted.
Most of the public-space community folks I've talked to around the country are not seeing anything like this kind of behavior, partly because community networks often (but not always) have enough people using them and aware of what's going on that it deters spammers and crackers. They're too exposed. They'd rather find unsecured mail servers than unsecured networks. They don't have a problem getting wireline access -- wireless isn't a big boon.
I mean, a cable modem in your house and some anonymous accounts work much better than freezing your tush in a park.
That said, there are a few simple things that could aid in making anyone trying to misuse a wireless connection, such as requiring a captive portal page a la nocat.net -- it scares people off. You can also use bandwidth restriction to make people don't abuse a segment. Sputnik.com has a free gateway software package that offers that option.
It's not about distance: it's about coverage. If you run point to point over a short distance, the signal becomes highly focused. These guys are not talking about boosting distances, but rather about increasing coverage without increasing ERP inside buildings, as one example.
If you had a 10,000 square foot facility and could put several access points (let's say 3 to 6) in a single server room with one of these antennas, you'd save a small fortune over, say, 20 to 30 access points all distributed across the actual area.
We'll see how it pans out.
This makes perfect sense. If you had 20 clients on an AP with full bandwidth saturation (bad network designer, bad!), and 15 of them had good IVs as did the AP, and 5 had bad, then the amount of traffic it takes is now proportionally higher based on how many packets have good IVs attached. if I understood your explanation correctly.
Thanks, much!
Does anyone know which of the newer firmware uses improved IVs? As I understand it, you can use an AP with better IVs without having to have a wireless adapter with any changes in firmware. A few companies, mostly in the enterprise space, have made announcements about their changes, but I haven't seen anything that isn't proprietary between client/AP from a single vendor.
No, the WET11 can handle up to 30 devices and bridge them to any access point. The WAP11, when turned into client mode, replicates a single MAC address as a single-device bridge. If you've figured out how to bridge multiple devices with a WAP11 to any access point, call the newspapers!
You know a lot of the terminology, but I'm having trouble believing you read the article or looked at what the product does.
The WET11 can be used for a single device (hence the ad hoc mode option) or to bridge up to 30 devices to an AP wirelessly.
If you've got a bunch of wired devices plugged into a switch and want to connect to another wired network somewhere else that already has an AP attached, you can use this to bridge those devices onto the other network.
Very simple idea, very simple product. Most of what you're complaining isn't something the article does.
But where are you sniffing? You can't associate with the WET11 and sit actively on the network. You could sniff traffic going by, but the WET11 is likely to be used in short-range point-to-point installations with some sort of antenna unless it's entirely within someone's home.
I don't buy "WEP is crackable in a few days." That's a canard. It's absolutely crackable with sufficient data samples. From my reading and talking with security folks, home networks don't generate the amount of traffic necessary over short periods of time, like days, and corporate IT managers should be boiled alive if they're letting non-encrypted data pass over wireless links.
So we agree!