Keep in mind that the IPv6 address space is more like 64 bits in many ways than 128 bits, since the smallest size of a network should be 64.
The goal was for this to be the last IP version needed, period.By the time IPv6 runs out (if it ever does), something will have replaced IP, just like IP replaced IPX.
Last time I tried DD-WRT, getting IPv6 on there was a CLI bitch, and I don't think it supported PD at that time.
I'm using pfSense now & never looked back to a SOHO router again. If you have an old P4 lying around with 512 ram, throw an extra 10/100 NIC in there & give it a spin. If you like it, you can roll your own fanless case & get the power consumption back down to an appliance level.
For a while I used the linksys I had as an access point, then swapped it out for a UniFi & again, couldn't be happier for the price.
Back to the topic, you'll find that the current state of IPv6 is not only an ISP issue, but also a hardware & software issue. Even pfSense only recently really supported IPv6 properly with 2.1, and many other devices I've tried have varying levels of support. Ironically, the most IPv6-complete item I've found is Windows 7/Server 2008R2.
While I'm on a kick recommending stuff, check out ipvfoo for Chrome once you get IPv6 working. It is helpful to see how many sites still don't support IPv6 native: https://code.google.com/p/ipvf...
Your WAN interface might be on a/128, and that is fine. You need to make sure your gear is telling Comcast what size of prefix you want delegated to your router.
Of course, this varies by market, so it might really not be there yet, but read up on prefix delegation & make sure you've got your end setup correctly: http://en.wikipedia.org/wiki/P...
Also, don't trust the tech support with this. They are clueless. According to them, IPv6 isn't available in my market.
and the potential is high for those 10% to 'assume' things are safe instead of driving their car like they're responsible for it.
Ok, I get what is going on. I'm assuming that the benefit is to self-driving cars, and I still think I'm right about that. More sensors mean better data for a computer to make better decisions.
Your point is that to a HUMAN driver, nearly all cars need to have it in order to provide useful information that human can react to. I'll agree with you there.
They won't be able to make all cars retroactively do this, and unless all cars are using this the system is essentially worthless.
Woah, worthless?
You're telling me that if there are 2 self-driving cars on the road with 20 old non-network cars, there is no benefit for those 2 cars to coordinate?
What about the two smartcars coordinating their shared view of all of the "dumb" cars? "Car 2, this is car 1. Dumbcar 249234 is next to me, in front of dumbtruck 2352 and has decreased speed from 72mph to 30mph and dropping." "ACK Car 1, dumbtruck 2352 is directly in front of me. Reducing speed to increase space between myself and dumbtruck 2352, preparing to change to lane 3 when available" *Obviously, they wouldn't talk like this, but could assign tags, coordinates & speeds, and this conversation could occur in milliseconds.
The idea is that the more sensors you have, the better the system works. Google and others are doing this with a single moving sensor platform now. A second one just makes both of them work better together, treating the non-connected cars as moving objects that must be avoided.
Looking at it another way, Waze is quite effective even with VERY low penetration rates. Imagine how much more effective Waze would be if every car made after 2018 was automatically plugged into it?
If you set the VoIP to low bandwidth requirements (sometimes erroneously marked as low quality), it'll be almost exactly like a POTS line -low latency and low fidelity.
Don't confuse VoIP over the Internet with VoIP over a managed network. You can run G.722 and have high fidelity, and run with 10ms packets and have low latency.
This is voice , not music, so you don't want hi-fi. A restricted frequency range actually makes voice much MORE intelligible because 95% of the intelligibility is in a narrow frequency range. The high and low frequencies are where the unwanted noise is.
Incorrect. High bandwidth audio makes many parts of speech more intelligible. For example, over quality VoIP, I can clearly hear the difference between M and N, where over POTS it can be much harder.
Plus, using the POTS range means your only hearing the harmonics of the real voice. I can't get where you think this would be better than hearing the full voice.
the government shouldn't be giving any money to any corporate interests
*IF* this bill actually did what you say here, I doubt many would be outraged. Yes, it would interfere with some Internet services, but it would also mean that the incumbents must start paying the city back for all right of way they are using at market rates.
It would also mean no more city or state money for the NFL stadiums, tax abatements for any business, etc.
I'm pretty sure most here could get on board with that.
Why replace something that is still as functional as the day it was made?
Because it makes the devices I bought yesterday far slower than they are designed to be.
If I have a 802.11b print server on my network, it might work fine. However, when I get home with my new 802.11n laptop & want to get on the web at 50mbit, that obsolete device can slow down my Netflix streaming because it hogs the channel for longer while someone prints to it.
More to the point, a single user in a public Wifi area (stadium, coffee house, etc) with 802.11b would cause EVERYONE to have a slower connection. Their device is now obsolete and should not be permitted on the network.
I think the point is that Cisco would like to ship their products with the slower stuff off, but if they do, they are no longer "Wi-Fi" compliant.
They're asking for a second "Wi-Fi" standard created so they can give the user a faster access point right out of the box & still be compliant with a standard.
I use Cisco wireless at work and Ubiquity at home. I have to say that there is still value for the Cisco products in larger companies.
The Ubnt stuff works OK at home, but there is no way I'd deploy a factory full of them using that java "controller" compared to Cisco's WLCs.
If you're a small business, sure, Ubnt is fine. If you have 300 sites to manage, you want something that can allow a single person to manage all of those networks from one console. The lower headcount can buy a LOT of expensive hardware.
So, I'm not following you on Slashdot, but this is the second time I've seen your link to Sue's page. I have to say, I love the way you continue to tell the world what a special woman she was.
Ever heard of the physics phenomenon where radiation decreases by the square of distance?
Yes, I'm fully aware of the inverse square law.
Wireless power needs to be within a few inches of the power source to work properly.
Well, good for wireless power, but that has absolutely nothing to do with 802.3at, which is a form of power over ethernet.
Ever heard of the 802.3 and how it is different from 802.11? Not every IEEE 802 standard has to do with wireless.
My point is that almost every laptop has an ethernet port. Use that port for both ethernet and power, which reduces the number of ports required and takes advantage of an existing worldwide DC power distribution standard: PoE.
Seriously, we can do 25.5W on 802.at-2009 NOW. Some vendors are doing 51W by using all 4 pairs.
Yes, I know many of you have laptops that draw almost 200 watts, but most of us don't need over 50W most of the time. If properly designed, the laptop can just "tread water" by slowing or stopping battery drain while drawing 51W during a work session, and then recharge while you're eating lunch or surfing Slashdot.
Imagine hooking your laptop up to power and ethernet at the same time! Single connection, less real estate used up on the exterior.
Just configure the laptop to draw power over the ethernet port, and not only do you not have to worry about a AC to DC brick, but you can travel the world and not have to worry about all the forms of AC power.
So, we agree that their 54Tbps number is total bullshit.:)
I was giving the marketing BS just to be generous. Of course, you will never really see 900mbit on every AP, or even a single AP, for all the reasons you mention and because they will never have a perfect 50/50 split of clients that are all evenly spread out among all of the access points.
I'm on your side. They claim 54Tbps, I say that 1.8Tbps is the *theoretical* max, and the practical Internet limit is 20Gbps, and now your talking about the real-world estimates, which will drop that 1.8T number to something much lower.
I noticed that, too. Assuming both 5Ghz and 2.4 Ghz radios at 450mbit each, you have 900 mbit per AP, X 2000, so 1.8 Tbit.
That, however, won't be the bottleneck. TFA says all of this has to go through redundant 10Gbit Internet connections. Maybe that means 20Gbit, and maybe a lot of the content will be local.
My bet, though, is that the bottleneck will be the Internet pipe if any slowness is detected. Depending on how many people live-stream video or have their devices set to auto-upload pictures and video to box/skydrive/gdrive/etc, I think they'll fill it.
First, thanks for sharing the technical details. You're among friends.
Second, I see this level of latency as a feature, not a bug. Much below 250ms and you'll have VoIP. At this level, email sync, file transfers (music and video) work fine, but RTC won't.
This is EXACTLY what I want on a flight. Either the ability to watch streaming movies & listen to music with my headphones on, or talk quietly to a neighbor while doing work on the laptop and NOT hearing everyone else have phone calls.
The change logs are for internal use; they aren't for your customers.
I'm not going to be one of your customers, then. I check the changelogs of every Cisco IOS update I deploy. If they changelog isn't published yet, I don't deploy.
I'm not going to add risk to my environment if all the new version did was fix bugs in features I don't use and add new features (with new bugs) that I won't use.
Slashdot Mirror
Keep in mind that the IPv6 address space is more like 64 bits in many ways than 128 bits, since the smallest size of a network should be 64.
The goal was for this to be the last IP version needed, period.By the time IPv6 runs out (if it ever does), something will have replaced IP, just like IP replaced IPX.
Last time I tried DD-WRT, getting IPv6 on there was a CLI bitch, and I don't think it supported PD at that time.
I'm using pfSense now & never looked back to a SOHO router again. If you have an old P4 lying around with 512 ram, throw an extra 10/100 NIC in there & give it a spin. If you like it, you can roll your own fanless case & get the power consumption back down to an appliance level.
For a while I used the linksys I had as an access point, then swapped it out for a UniFi & again, couldn't be happier for the price.
Back to the topic, you'll find that the current state of IPv6 is not only an ISP issue, but also a hardware & software issue. Even pfSense only recently really supported IPv6 properly with 2.1, and many other devices I've tried have varying levels of support.
Ironically, the most IPv6-complete item I've found is Windows 7/Server 2008R2.
While I'm on a kick recommending stuff, check out ipvfoo for Chrome once you get IPv6 working. It is helpful to see how many sites still don't support IPv6 native: https://code.google.com/p/ipvf...
I'm on Comcast, and I'm getting a /60 from them.
Your WAN interface might be on a /128, and that is fine. You need to make sure your gear is telling Comcast what size of prefix you want delegated to your router.
Of course, this varies by market, so it might really not be there yet, but read up on prefix delegation & make sure you've got your end setup correctly: http://en.wikipedia.org/wiki/P...
Also, don't trust the tech support with this. They are clueless. According to them, IPv6 isn't available in my market.
Exactly this.
People care more about their favored video service (Netflix, Amazon, iTunes) than their ISP.
The content providers need to show how the ISPs affect the speed, and who the best option in your area is.
and the potential is high for those 10% to 'assume' things are safe instead of driving their car like they're responsible for it.
Ok, I get what is going on. I'm assuming that the benefit is to self-driving cars, and I still think I'm right about that. More sensors mean better data for a computer to make better decisions.
Your point is that to a HUMAN driver, nearly all cars need to have it in order to provide useful information that human can react to. I'll agree with you there.
They won't be able to make all cars retroactively do this, and unless all cars are using this the system is essentially worthless.
Woah, worthless?
You're telling me that if there are 2 self-driving cars on the road with 20 old non-network cars, there is no benefit for those 2 cars to coordinate?
What about the two smartcars coordinating their shared view of all of the "dumb" cars?
"Car 2, this is car 1. Dumbcar 249234 is next to me, in front of dumbtruck 2352 and has decreased speed from 72mph to 30mph and dropping."
"ACK Car 1, dumbtruck 2352 is directly in front of me. Reducing speed to increase space between myself and dumbtruck 2352, preparing to change to lane 3 when available"
*Obviously, they wouldn't talk like this, but could assign tags, coordinates & speeds, and this conversation could occur in milliseconds.
The idea is that the more sensors you have, the better the system works. Google and others are doing this with a single moving sensor platform now. A second one just makes both of them work better together, treating the non-connected cars as moving objects that must be avoided.
Looking at it another way, Waze is quite effective even with VERY low penetration rates. Imagine how much more effective Waze would be if every car made after 2018 was automatically plugged into it?
Lol! You're telling me that better than listening to someone in person I should record them & put it through a filter?
Sure. If there is background noise, I get it. Two people talking in a quiet room? No way.
If you set the VoIP to low bandwidth requirements (sometimes erroneously marked as low quality), it'll be almost exactly like a POTS line -low latency and low fidelity.
Don't confuse VoIP over the Internet with VoIP over a managed network. You can run G.722 and have high fidelity, and run with 10ms packets and have low latency.
This is voice , not music, so you don't want hi-fi. A restricted frequency range actually makes voice much MORE intelligible because 95% of the intelligibility is in a narrow frequency range. The high and low frequencies are where the unwanted noise is.
Incorrect. High bandwidth audio makes many parts of speech more intelligible. For example, over quality VoIP, I can clearly hear the difference between M and N, where over POTS it can be much harder.
Plus, using the POTS range means your only hearing the harmonics of the real voice. I can't get where you think this would be better than hearing the full voice.
the government shouldn't be giving any money to any corporate interests
*IF* this bill actually did what you say here, I doubt many would be outraged. Yes, it would interfere with some Internet services, but it would also mean that the incumbents must start paying the city back for all right of way they are using at market rates.
It would also mean no more city or state money for the NFL stadiums, tax abatements for any business, etc.
I'm pretty sure most here could get on board with that.
There is no way the gov. authorities could have prevented the problem
Try again:
http://www.wunderground.com/ne...
All they had to do was cancel school that day.
Your point is complete correct & I was aware & agree with you.
However, my point was to the GP who was saying that nothing is obsolete. Hopefully, between the two of us, he will apologize for being so wrong. ;)
Why replace something that is still as functional as the day it was made?
Because it makes the devices I bought yesterday far slower than they are designed to be.
If I have a 802.11b print server on my network, it might work fine. However, when I get home with my new 802.11n laptop & want to get on the web at 50mbit, that obsolete device can slow down my Netflix streaming because it hogs the channel for longer while someone prints to it.
More to the point, a single user in a public Wifi area (stadium, coffee house, etc) with 802.11b would cause EVERYONE to have a slower connection. Their device is now obsolete and should not be permitted on the network.
I think the point is that Cisco would like to ship their products with the slower stuff off, but if they do, they are no longer "Wi-Fi" compliant.
They're asking for a second "Wi-Fi" standard created so they can give the user a faster access point right out of the box & still be compliant with a standard.
I use Cisco wireless at work and Ubiquity at home. I have to say that there is still value for the Cisco products in larger companies.
The Ubnt stuff works OK at home, but there is no way I'd deploy a factory full of them using that java "controller" compared to Cisco's WLCs.
If you're a small business, sure, Ubnt is fine. If you have 300 sites to manage, you want something that can allow a single person to manage all of those networks from one console. The lower headcount can buy a LOT of expensive hardware.
"Below zero" is a turn of phrase that means freezing to 96% of the world's population. It is used in the United States to mean below 0F.
FTFY
So, I'm not following you on Slashdot, but this is the second time I've seen your link to Sue's page.
I have to say, I love the way you continue to tell the world what a special woman she was.
Have you considered using DHCP reservations?
Ever heard of the physics phenomenon where radiation decreases by the square of distance?
Yes, I'm fully aware of the inverse square law.
Wireless power needs to be within a few inches of the power source to work properly.
Well, good for wireless power, but that has absolutely nothing to do with 802.3at, which is a form of power over ethernet.
Ever heard of the 802.3 and how it is different from 802.11? Not every IEEE 802 standard has to do with wireless.
My point is that almost every laptop has an ethernet port. Use that port for both ethernet and power, which reduces the number of ports required and takes advantage of an existing worldwide DC power distribution standard: PoE.
I deploy PoE switches every week.
If you're using a proper switch & power supplies, there is no issue with running PoE+ on every port of the switch.
If you're using some bargain SOHO switch, yes, you might only be able to power 1/3 of the ports a full PoE+ or 1/2 of them at PoE.
An unladen honeybee is about 80mg.
Is that an African or European honeybee?
Seriously, we can do 25.5W on 802.at-2009 NOW. Some vendors are doing 51W by using all 4 pairs.
Yes, I know many of you have laptops that draw almost 200 watts, but most of us don't need over 50W most of the time. If properly designed, the laptop can just "tread water" by slowing or stopping battery drain while drawing 51W during a work session, and then recharge while you're eating lunch or surfing Slashdot.
Imagine hooking your laptop up to power and ethernet at the same time! Single connection, less real estate used up on the exterior.
Just configure the laptop to draw power over the ethernet port, and not only do you not have to worry about a AC to DC brick, but you can travel the world and not have to worry about all the forms of AC power.
So, we agree that their 54Tbps number is total bullshit. :)
I was giving the marketing BS just to be generous. Of course, you will never really see 900mbit on every AP, or even a single AP, for all the reasons you mention and because they will never have a perfect 50/50 split of clients that are all evenly spread out among all of the access points.
I'm on your side. They claim 54Tbps, I say that 1.8Tbps is the *theoretical* max, and the practical Internet limit is 20Gbps, and now your talking about the real-world estimates, which will drop that 1.8T number to something much lower.
In any case, 54Tbps = bullshit.
I noticed that, too. Assuming both 5Ghz and 2.4 Ghz radios at 450mbit each, you have 900 mbit per AP, X 2000, so 1.8 Tbit.
That, however, won't be the bottleneck. TFA says all of this has to go through redundant 10Gbit Internet connections. Maybe that means 20Gbit, and maybe a lot of the content will be local.
My bet, though, is that the bottleneck will be the Internet pipe if any slowness is detected. Depending on how many people live-stream video or have their devices set to auto-upload pictures and video to box/skydrive/gdrive/etc, I think they'll fill it.
First, thanks for sharing the technical details. You're among friends.
Second, I see this level of latency as a feature, not a bug. Much below 250ms and you'll have VoIP. At this level, email sync, file transfers (music and video) work fine, but RTC won't.
This is EXACTLY what I want on a flight. Either the ability to watch streaming movies & listen to music with my headphones on, or talk quietly to a neighbor while doing work on the laptop and NOT hearing everyone else have phone calls.
The change logs are for internal use; they aren't for your customers.
I'm not going to be one of your customers, then. I check the changelogs of every Cisco IOS update I deploy. If they changelog isn't published yet, I don't deploy.
I'm not going to add risk to my environment if all the new version did was fix bugs in features I don't use and add new features (with new bugs) that I won't use.