11n FUD? How interesting. The manufactures of 11n devices make up the majority of the standards committee. They aren't going to add new hardware requirements to the standard. Finalization is simply a matter of bureaucracy. Meanwhile, have you noticed how all the devices interoperate now? Did you spot the Wi-Fi certifications? It's already a working standard.
Do you honestly expect your AP to stop working with your laptop when they file the final papers? Will the backward compatibility with 11g and 11a disappear? No. And FYI, the final standard standard will work with draft 2.0 too. Your concerns were valid with pre-N equipment, but now it's unfounded.
Actually, they advertise 11n draft 2.0 these days, not pre-11n. There's enough draft 2.0 devices around to choose from. If you keep worrying about future standards, you'll never buy anything.
Wouldn't that be overkill? All 11n devices do MIMO currently - if a future device does not, it won't have a > 200Mbps number on the box. Greenfield is insignificant in practice. You just want to look out for 2.4 and 5 GHz support, or dual radio if you need that. I realize that retailers try to fool people by putting just a giant 11N on the box, but if you read closer you'll see if 5 Ghz, or 11a compatibility, is included.
As far as maintaining a website, most of these products come and go too fast to keep up with. The user could write down model XGY-2000T and find a XGY-2001B when they get to the store. Better to just read the small print on the box.
B and G use the same channels, and 11n can use G or A channels. So, 'NA' is open too. Another advantage of the A channels is that there is less channel overlap and a shorter range. In a small house, 11a will still cover it all, but neighboring houses will become weaker interferers. In a condo situation, 11a is ideal.
My second choice if it's a single user, and only for internet browsing, would be 11b (or 11g limited to 11mbs rate - same thing). It really does seem to work better in practice, although I expected rate adaptation algorithms to find the best rate for my 11g connection, it is not so. 5-7 Mbps should be enough for this user anyway.
Sadly, retailers have been selling us b/g-only for a long time to save a couple of bucks, and nobody seems to care.
"i love how intel is touting how (even though wifi has been tech since the late 90's early 2000's) that they finally got around to making it work over "long" distances."
I love that they chose to make a big deal about a Wi-Fi solution when WiMax was supposed to be here already. Wasn't it this time last year when they were talking about WiMax adapters being standard on laptops?...and over long distances too?! This the kind of tech that Intel should kill, not promote... if WiMax were really on its way, that is.
With N you're possibly doubling the number of radios to four.
Ouch, that's embarrassing. Maybe you should've prefaced your comments with, "I'm only guessing here, but..."
Still two radios for simultaneous 11n in 2.4 GHz and 5 Ghz. Noise floor is likewise not the problem you think it is. Instead, we have channel bonding, which means that you're transmitting across a broader spectrum, and multiple transmit chains (MIMO) means that your transmits are less efficient, even at the same power level as non-11n.
BTW, I'm not saying the radio power increase is not significant, especially the full 3-by-3 implementation that Siemens claims to have used, which made it more remarkable.
If it was tons of power, how could they make 802.11n adapters for laptops?
A couple of points:
1) Laptop clients are single-radios, not the two radios that the dual-band access points have.
2) All the other components that the access point needs to have is already powered by the laptop, e.g. cpu, memory, ethernet switch. The client, being just one radio, can draw less than 2 W during continuous 11n transmit.
Bonus point: The client can also sleep the radio when not in use. The AP has to stay awake and beaconing all the time.
Since the Siemens access point design uses a standard 802.11n radio, clearly their power savings were made on the rest of the board.
It's gonna be hard to offer security with no user interface on the camera, and I wouldn't use it without that. Once most cameras offer built-in Wi-Fi, these little gadgets - although cool - will be overpriced and obsolete.
This issue has nothing to do with non-free. The fix was in the free portion of the code, according to this link. Moreover, the bug was not found due to it being open source. It was found by testing. So just by virtue of being FOSS, software does not become miraculously bug-free.
For the driver to be completely free, Atheros would have to create a chip incapable of violating FCC regulations, but then it couldn't be sold in Europe, because it wouldn't support all the channels available there. Global OEMs wouldn't touch it. I think Atheros is being as helpful as they can without having their product pulled from the market by regulatory authorities. The video chip makers have the option to support fully open drivers. The wireless chip makers do not.
a small open-source Linux compatibility shim around the actual, binary only driver.
So the binary HAL layer is less than half of my driver and doesn't include frame parsing and generation or rate control, yet you'd like to call it a small compatibility shim? I'd say the driver is mostly open source.
As for the effort to reverse engineer the HAL, I think the chip versions are revised too quickly for that to be widely successful. Seems like a lot of work for little return.
That phrase has always bothered me. The absence of evidence, assuming one has looked, *is* evidence of absence. It's just not proof of absence. In the same way that circumstantial evidence is still evidence although it may not be conclusive. We use this in science all the time when we look for some evidence that we expect to be there. When it's absent, that tells us something.
So, this isn't good enough for you? Ok, but it's customary to say why not. I think that particular cliché does us a disservice.
Of course, once you do that, it becomes even more obvious that the content to garbage ratio on that site is well below 50%, but at least you don't have to click and wait at each break.
Maybe it's because Jobs declared 802.11a dead a few years ago and nobody wants to make him look bad. I mean how could it possibly hurt sales to tell people it supports additional channels that they might use in the future?
I think George Ou brings up 3rd party drivers as a distraction and because it's an easier position to defend. Unfortunately for him, this isn't about 3rd party drivers for two reasons. First, David Maynor has admitted that the presentation given to Brian Krebs, before the "Hijack a Mac in 60 Seconds" piece that started it all, did not use a 3rd party wireless card. Secondly, he hasn't released the 3rd party exploit either!!
Maynor is responsible for the media attention, and Apple's response. Of course, all of that would mean nothing, and he would be a superstar hacker if he just released his exploit. He could do it for a clueless reporter on demand in August of last year. Now, eight months later, it's too hard to reproduce in front of a technical audience? Sounds like a rigged demo to me.
Apple isn't as friendly or responsive to security researchers as they should be from what I can tell, but none of that is an issue given the magnitude of Maynor and Ellch's misconduct.
The 2.0 draft isn't really guaranteed either, if by that you mean that they will take financial responsibility if your particular hardware cannot be upgraded to 100% compliance. I read "guarantee" in a different news item as well, but I think it just means that they don't plan to accept any more changes that could possibly require a hardware change to support - as far as they know. Of course, your particular firmware might not be architected right, and although it could support the next round of minor changes, that doesn't mean your manufacturer will bother to update your firmware - hey, it should mostly work.
I don't see much difference between the 1.0 draft. We pretty much knew that that hardware would support final draft, since there was a partnership outside the IEEE process before it was submitted, but there was a lot of useless pondering by tech journalists trying to suggest otherwise and appear "in the know". In reality, the situation hasn't changed much since 1.0.
Nevertheless, I'm in no rush to buy it yet. Remember, if it doesn't have some sort of certification on it, that means it's working in the manufacturer's opinion. Good luck proving otherwise.
P.S. Almost nobody guaranteed upgradability of the 1.0 stuff. I don't know why. I suppose it was just too big a number on the liability sheet, regardless of how confident they were. I only saw ASUS make a guarantee, and that was for the last 3 months of 2006 - subsequently withdrawn. Pretty lame, huh?
They're already facing the immediate punishment of a car accident in which they might die. I doubt that yet another overpriced moving violation will do any good.
What we need is a good mobile RF jammer. Now there's a remedy that any citizen can enforce.
"- remove wireless card (save $10)" Also onboard, no can do.
No it isn't. Even a centrino laptop has a miniPCI or a PCIx card in it. Other than your facts, I don't think you're so wrong.
I think high volumes sales will require a finished product from Dell. Obviously they would have to pre-install something and have it 100% working.
I also think the OP has the wrong idea about options. They should just require their existing vendors to provide Linux drivers. They won't have to pay more, because vendors will fall over themselves to protect that revenue stream, but we will have to wait until the next generation platforms are ready. That's OK, we have to wait until Dell figures out the best way to offer support anyway.
In the end, the upgrades option are a way to get buyers to pay more for their product and still feel good about their purchase. It's about price points just as much as meeting customer needs.
I guess Novell prefers to make vague statements about an "uptick" in Linux sales than share an example of a US firm widely adopting Linux on the desktop. It would be those kinds of examples that would make it harder for people to make ignorant statements about Microsoft saving them millions of dollars. If Linux doesn't suit your needs, just say so, but don't lie about how much money it's saving you.
Slashdot Mirror
11n FUD? How interesting. The manufactures of 11n devices make up the majority of the standards committee. They aren't going to add new hardware requirements to the standard. Finalization is simply a matter of bureaucracy. Meanwhile, have you noticed how all the devices interoperate now? Did you spot the Wi-Fi certifications? It's already a working standard.
Do you honestly expect your AP to stop working with your laptop when they file the final papers? Will the backward compatibility with 11g and 11a disappear? No. And FYI, the final standard standard will work with draft 2.0 too. Your concerns were valid with pre-N equipment, but now it's unfounded.
Actually, they advertise 11n draft 2.0 these days, not pre-11n. There's enough draft 2.0 devices around to choose from. If you keep worrying about future standards, you'll never buy anything.
Wouldn't that be overkill? All 11n devices do MIMO currently - if a future device does not, it won't have a > 200Mbps number on the box. Greenfield is insignificant in practice. You just want to look out for 2.4 and 5 GHz support, or dual radio if you need that. I realize that retailers try to fool people by putting just a giant 11N on the box, but if you read closer you'll see if 5 Ghz, or 11a compatibility, is included.
As far as maintaining a website, most of these products come and go too fast to keep up with. The user could write down model XGY-2000T and find a XGY-2001B when they get to the store. Better to just read the small print on the box.
B and G use the same channels, and 11n can use G or A channels. So, 'NA' is open too. Another advantage of the A channels is that there is less channel overlap and a shorter range. In a small house, 11a will still cover it all, but neighboring houses will become weaker interferers. In a condo situation, 11a is ideal.
My second choice if it's a single user, and only for internet browsing, would be 11b (or 11g limited to 11mbs rate - same thing). It really does seem to work better in practice, although I expected rate adaptation algorithms to find the best rate for my 11g connection, it is not so. 5-7 Mbps should be enough for this user anyway.
Sadly, retailers have been selling us b/g-only for a long time to save a couple of bucks, and nobody seems to care.
"i love how intel is touting how (even though wifi has been tech since the late 90's early 2000's) that they finally got around to making it work over "long" distances."
...and over long distances too?! This the kind of tech that Intel should kill, not promote... if WiMax were really on its way, that is.
I love that they chose to make a big deal about a Wi-Fi solution when WiMax was supposed to be here already. Wasn't it this time last year when they were talking about WiMax adapters being standard on laptops?
With N you're possibly doubling the number of radios to four.
Ouch, that's embarrassing. Maybe you should've prefaced your comments with, "I'm only guessing here, but..."
Still two radios for simultaneous 11n in 2.4 GHz and 5 Ghz. Noise floor is likewise not the problem you think it is. Instead, we have channel bonding, which means that you're transmitting across a broader spectrum, and multiple transmit chains (MIMO) means that your transmits are less efficient, even at the same power level as non-11n.
BTW, I'm not saying the radio power increase is not significant, especially the full 3-by-3 implementation that Siemens claims to have used, which made it more remarkable.
If it was tons of power, how could they make 802.11n adapters for laptops?
A couple of points:
1) Laptop clients are single-radios, not the two radios that the dual-band access points have.
2) All the other components that the access point needs to have is already powered by the laptop, e.g. cpu, memory, ethernet switch. The client, being just one radio, can draw less than 2 W during continuous 11n transmit.
Bonus point: The client can also sleep the radio when not in use. The AP has to stay awake and beaconing all the time.
Since the Siemens access point design uses a standard 802.11n radio, clearly their power savings were made on the rest of the board.
Our 802.11a/b/g access points here currently draw about 4.5 W each
Check again, it's probably more like 12 W. Or post the model number of this dual band wonder.
Not much different than this Wi-Fi SD card
It's gonna be hard to offer security with no user interface on the camera, and I wouldn't use it without that. Once most cameras offer built-in Wi-Fi, these little gadgets - although cool - will be overpriced and obsolete.
At least Xandros can't say they didn't see community backlash and GPLv3 coming. Here it comes...
Only one decade old? It looks a lot older. He's had a good run, but it's time we updated it? Does everyone agree?
you can type, "lsmod | grep ath_pci" to find out if you are running the supposedly exploited module
You can also type "modinfo ath_pci | grep version" to find which version you have.
The patched driver is 0.9.2.1 or newer.
This issue has nothing to do with non-free. The fix was in the free portion of the code, according to this link. Moreover, the bug was not found due to it being open source. It was found by testing. So just by virtue of being FOSS, software does not become miraculously bug-free.
For the driver to be completely free, Atheros would have to create a chip incapable of violating FCC regulations, but then it couldn't be sold in Europe, because it wouldn't support all the channels available there. Global OEMs wouldn't touch it. I think Atheros is being as helpful as they can without having their product pulled from the market by regulatory authorities. The video chip makers have the option to support fully open drivers. The wireless chip makers do not.
a small open-source Linux compatibility shim around the actual, binary only driver.
So the binary HAL layer is less than half of my driver and doesn't include frame parsing and generation or rate control, yet you'd like to call it a small compatibility shim? I'd say the driver is mostly open source.
As for the effort to reverse engineer the HAL, I think the chip versions are revised too quickly for that to be widely successful. Seems like a lot of work for little return.
"the absence of evidence..."
That phrase has always bothered me. The absence of evidence, assuming one has looked, *is* evidence of absence. It's just not proof of absence. In the same way that circumstantial evidence is still evidence although it may not be conclusive. We use this in science all the time when we look for some evidence that we expect to be there. When it's absent, that tells us something.
So, this isn't good enough for you? Ok, but it's customary to say why not. I think that particular cliché does us a disservice.
You're breaking the social contract, you bastard!
So, would it be wrong to show people the whole list, allowing many people to ignore their ad-laden web page altogether?
Fasterfox
NoScript
Adblock Plus
PDF Download
VideoDownloader
Greasemonkey
ScribeFire
TrackMeNot
Tabbrowser Preferences
Tabbrowser Extensions
FormSpy
Hmmm. It doesn't feel wrong.
gives me an idea for another extension which stitches these kinds of articles together
Firefox repagination: http://addons.mozilla.org/firefox/2099
Of course, once you do that, it becomes even more obvious that the content to garbage ratio on that site is well below 50%, but at least you don't have to click and wait at each break.
"With an Apple product, "no word" definitely means it doesn't play them."
Counterexample: Did you ever notice how the MacBook specs never tell you it does 802.11a? 802.11g standard; 802.11n capable
Maybe it's because Jobs declared 802.11a dead a few years ago and nobody wants to make him look bad. I mean how could it possibly hurt sales to tell people it supports additional channels that they might use in the future?
I think George Ou brings up 3rd party drivers as a distraction and because it's an easier position to defend. Unfortunately for him, this isn't about 3rd party drivers for two reasons. First, David Maynor has admitted that the presentation given to Brian Krebs, before the "Hijack a Mac in 60 Seconds" piece that started it all, did not use a 3rd party wireless card. Secondly, he hasn't released the 3rd party exploit either!!
Maynor is responsible for the media attention, and Apple's response. Of course, all of that would mean nothing, and he would be a superstar hacker if he just released his exploit. He could do it for a clueless reporter on demand in August of last year. Now, eight months later, it's too hard to reproduce in front of a technical audience? Sounds like a rigged demo to me.
Apple isn't as friendly or responsive to security researchers as they should be from what I can tell, but none of that is an issue given the magnitude of Maynor and Ellch's misconduct.
don't see why this is on the front page. This is a question to the public, not news.
/.?
Apparently you're not familiar with the AskSlashdot section. I'll leave it to you to discover what kinds of articles are posted there.
Secondly, the summary was not posted to the main page, it is just linked there as all articles are now.
Take it to a freakin forum.
You mean like
The 2.0 draft isn't really guaranteed either, if by that you mean that they will take financial responsibility if your particular hardware cannot be upgraded to 100% compliance. I read "guarantee" in a different news item as well, but I think it just means that they don't plan to accept any more changes that could possibly require a hardware change to support - as far as they know. Of course, your particular firmware might not be architected right, and although it could support the next round of minor changes, that doesn't mean your manufacturer will bother to update your firmware - hey, it should mostly work.
I don't see much difference between the 1.0 draft. We pretty much knew that that hardware would support final draft, since there was a partnership outside the IEEE process before it was submitted, but there was a lot of useless pondering by tech journalists trying to suggest otherwise and appear "in the know". In reality, the situation hasn't changed much since 1.0.
Nevertheless, I'm in no rush to buy it yet. Remember, if it doesn't have some sort of certification on it, that means it's working in the manufacturer's opinion. Good luck proving otherwise.
P.S. Almost nobody guaranteed upgradability of the 1.0 stuff. I don't know why. I suppose it was just too big a number on the liability sheet, regardless of how confident they were. I only saw ASUS make a guarantee, and that was for the last 3 months of 2006 - subsequently withdrawn. Pretty lame, huh?
They're already facing the immediate punishment of a car accident in which they might die. I doubt that yet another overpriced moving violation will do any good.
What we need is a good mobile RF jammer. Now there's a remedy that any citizen can enforce.
"- remove wireless card (save $10)"
Also onboard, no can do.
No it isn't. Even a centrino laptop has a miniPCI or a PCIx card in it. Other than your facts, I don't think you're so wrong.
I think high volumes sales will require a finished product from Dell. Obviously they would have to pre-install something and have it 100% working.
I also think the OP has the wrong idea about options. They should just require their existing vendors to provide Linux drivers. They won't have to pay more, because vendors will fall over themselves to protect that revenue stream, but we will have to wait until the next generation platforms are ready. That's OK, we have to wait until Dell figures out the best way to offer support anyway.
In the end, the upgrades option are a way to get buyers to pay more for their product and still feel good about their purchase. It's about price points just as much as meeting customer needs.
I guess Novell prefers to make vague statements about an "uptick" in Linux sales than share an example of a US firm widely adopting Linux on the desktop. It would be those kinds of examples that would make it harder for people to make ignorant statements about Microsoft saving them millions of dollars. If Linux doesn't suit your needs, just say so, but don't lie about how much money it's saving you.
We'll miss all of it here in California. The total eclipse happens at 2:45pm PST, and will last about an hour after that. Sunset is at 6pm.
We'll have our turn on Aug. 28, when the tables (and the planet) will be turned. Mark your calender.