Great article! If, however, you happen to be in Switzerland at any time during the next couple of months there is an Albert Einstein exhibit at the History Museum in Bern. There are audio tours given in all languages, and it covers both his scientific achievements and his personal life. The exhibit attempts to explain his achievements in a way that uninitiated would understand, and it succeeds somewhat. For more info see http://www.bhm.ch/en/ausstellungen_sonder_01.cfm
Why does any of this matter? Intel-based processors are going to support 802.11i no matter what the ITU has to say, and China is going to mandate whatever they're going to mandate, again no matter what the ITU has to say. Isn't this just a matter of face?
Isn't this the same Bob Metcalfe who said the Internet would be dead by 1996 or something like that? I think I have a NANOG tee-shirt that shows him eating his words.
The reporter in the article seems to think that people can easily protect themselves on wireless networks, and we all know that just isn't true for several reasons:
- Depending on the card you buy PCs sometimes have trouble converting ASCII to bits in the same way. I have this problem with, say my NETGEAR and my Mac.
- WEP sucks and we all know it, so 15 minutes of a determined script kiddie's time and that's the ball game.
- WPA isn't yet available everywhere, and even it is supposed to be an interim standard to 802.11i.
In short, you can only avoid nuisance freeloading with WEP and it's a pain to use if you have multiple PCs. Especially if you're not the sort that reads/..
Dan Farmer was asked to leave (then) large computer manufacturer when he released Satan. Dan Geer was asked to leave a so-called security company when he and other security notables argued in favor of heterogeneity for purposes of resilience. In the end, it was the companies that looked like idiots and not the above individuals. Bonus for the person who can name the companies...
traditional tel companies can lobby congress to drive up the costs because idiots don't realize 911 won't work, which was told to the purchaser prior to the sale!
It's really not that simple. For one thing, there has been a lot of work to make 911 service work, and in fact it will. There are several different standards out there, but probably the most robust one involves GEOPRIV (RFC 3825 amongst others).
But beyond that, were we to accept the above argument that VOIP vendors not provide for the same safety of individuals as traditional telcos we would all see our safety degrade as the E911 system loses its enonomies of scale. This isn't something people think about, either, until you really need it.
So given that we have a technical solution, it really doesn't cost that much to implement, and it's in society's interest to ease access to emergency services, what's the problem?
Test users' appetite for downloads?? Why don't they just ask Apple or an Apple shareholder? Better yet, put the old Hitchhikers Guide or Dr. Who episodes online and they'll find out right quick about appetites for downloads!
Create a new account on your favorite free email server, like Yahoo! or Hotmail! or whoever just for the vacation and give that to people in advance. Tell people not to send sensitive stuff, but if something important comes up to have messages like "Please call Elsa. It's important."
You can even leave a.vacation.msg that gives proper instructions on the use of te throw-away account.
This doesn't work for every use, but I could definitely understand the need to have some means of communication when one is on the go. This is especially true when going from country to country, say, in Europe, where I live.
In fact it was important for a friend who was bouncing between Germany and Switzerland who was having trouble connecting with a host. If you're going to be away so long that you need to do more extensive (and sensitive) stuff then I agree with the other posters that you should either bring a laptop or use smart technology. In fact, see my article about a proposed solution here.
On its own I wouldn't discount what Semantec says. However, "simulations" generally involve models, and those models have assumptions. What are the assumptions in this model, I wonder? We already know that a virus can travel roughly at the speed of a disk drive's ability to write.
It would be more interesting to see a study of computer-based virii versus biological ones. How about some real epidemiologists take a crack at it? Perhaps they already have..
While Londoners are using Microphones to handle noise violations, the people of East Palo Alto, California used them for accoustic analysis to determine from which direction the bullets were flying in a particularly gang-ridden neighborhood, and it was reported to have worked. Very few residents complained about their rights being violated.
Now, on the other hand, if London started recording conversations, that would make for a more interesting - and invasive- use. At that point we're heading right for 1984. Of course, London already uses cameras in public places for use as evidence, so their networking infrastructure is already there...
This report says absolutely nothing new. If you're going to take VoIP seriously, you need to recognize the application's needs. In this case, some amount of QoS is important, particularly at conjestion points such as the last hop to the consumer. You also need to recognize that like any other application on the Internet DDOS is a possibility. Ain't no different.
On the other hand, IPv6 will solve all our problems, right?;-)
Microsoft breaking things doesn't help upgrades, indeed. Those sorts of security fixes are tricky, and sometimes the inherent functionality in use is the bug. Indeed the Mozilla folks just dealt with a similar matter for Firefox 1.0.3. For an amusing set of conversations check out Bugzilla bugs 289231 and 281988 where they successfully grappled with the same sorts of problems.
Terrestrially speaking, that formula's not quite right because no man made signal I know of goes through the center of the earth. You have to go around. The best you could possibly do at the equator is pi R in each direction or 2 pi R, which is > 6r.
That's about 0.133 ms.
At that bound VoIP is still good. In general codecs are good enough such that with buffering you can get away with up to 200ms without a human noticing. Beyond 200ms people start tripping over each other because of delay.
The gentleman who runs the Iraqi ISP mentioned that he could get 500ms round trip times. At best the people on the ends of these phones should demonstrate patience with the technology.
... whatever yore means, there was one service provider based in Texas whose management were quite offended by dirty pictures on USENET. So they hired a consultant to see just how much money that would save on... disk space, yes disk space.
The contractor came back and said that they could save 60% of their disk space, but since he also analyzed their NNTP logs he told them that porn was also their major source of revenue.
They're still around. Guess which path they chose.
VoIP is still not a complete solution, at least not for reliable service just yet, IMHO
Drops are the result of networks that are not properly engineered for voice. You need to do that.
That means using appropriate QoS mechanisms at the edges and either using those same QoS mechanisms across the core or being sure to over-engineer the links (your mileage may vary).
VOIP is a tradeoff in this regard. Yes, you don't have to duplicate infrastructure, and you actually do get more efficient use of the bandwidth. In addition, the codecs used for IP phones arefar more resilient to both loss and delay, as compared to an analog phone. An analog phone has no buffers to play with.
Not all of ISOC's programs benefit all of dot-org. So why tax all of dot-org?
Since the vast majority of ISOC's programs involve the IETF and standards organization support, it's really hard for me to accept the above statement.
The ones that do benefit all of dot-org benefit more than just dot-org. So why tax just dot-org?
I have no objection to taxing other domains, but I don't see the mechanism necessary to do the job. Furthermore, I'm actually betting that ISOC can "tax".org and yet come in well below the costs of other domains.
As for the rest of ISOC's stuff, find the right people to charge.
That's easy to say and hard to do. I would argue that since we all benefit from ISOC we should all pay a small fee, and so again I would accept the notion that all domain owners pay a small fee.
I would be happy to continue this conversation. To do so, disentangle my email address and feel free to mail me.
Want to get your document through the IETF process? Well, the IETF is going to need to scale. In order to do that, they're probably going to need money. And where will that money come from? How about those of us who benefit from their standards? Are we talking beaceaup bucks? Probably not, but I could certainly think of worse places for the money to go.
If security of VOIP is no WORSE than that of a normal phone line, then at least you only have to worry about physically securing a single infrastructure as opposed to the multiple parallel ones you had to manage previously.
Also, in some environments VOIP on top of IPsec may be reasonable. The article is NOT entirely on target (IMHO it's a cheap hit piece). Consider the Cisco IP Phone in a work station. Since a person plugs his or her PC into the phone for network connectivity, you merely need to have some way to trust that the phone is authorized to use QoS, and you can thus encrypt voice traffic AND have the phone do classification.
And that trust can be gotten on the small with simple approaches such as MAC address lockdowns on your switches.
There was the incident of the fellow who discovered that the New York Times was left wide open by FrontPage. So he called to tell them, and was promptly arrested. I wonder if Mr. Clarke thinks that's fair.
Slashdot Mirror
Great article! If, however, you happen to be in Switzerland at any time during the next couple of months there is an Albert Einstein exhibit at the History Museum in Bern. There are audio tours given in all languages, and it covers both his scientific achievements and his personal life. The exhibit attempts to explain his achievements in a way that uninitiated would understand, and it succeeds somewhat. For more info see http://www.bhm.ch/en/ausstellungen_sonder_01.cfm
Why does any of this matter? Intel-based processors are going to support 802.11i no matter what the ITU has to say, and China is going to mandate whatever they're going to mandate, again no matter what the ITU has to say. Isn't this just a matter of face?
It is here.
Isn't this the same Bob Metcalfe who said the Internet would be dead by 1996 or something like that? I think I have a NANOG tee-shirt that shows him eating his words.
The reporter in the article seems to think that people can easily protect themselves on wireless networks, and we all know that just isn't true for several reasons:
/..
- Depending on the card you buy PCs sometimes have trouble converting ASCII to bits in the same way. I have this problem with, say my NETGEAR and my Mac.
- WEP sucks and we all know it, so 15 minutes of a determined script kiddie's time and that's the ball game.
- WPA isn't yet available everywhere, and even it is supposed to be an interim standard to 802.11i.
In short, you can only avoid nuisance freeloading with WEP and it's a pain to use if you have multiple PCs. Especially if you're not the sort that reads
half right. SGI and @stake. Although I'm sorry anyone remembers either, given their cowardace.
Dan Farmer was asked to leave (then) large computer manufacturer when he released Satan. Dan Geer was asked to leave a so-called security company when he and other security notables argued in favor of heterogeneity for purposes of resilience. In the end, it was the companies that looked like idiots and not the above individuals. Bonus for the person who can name the companies...
But beyond that, were we to accept the above argument that VOIP vendors not provide for the same safety of individuals as traditional telcos we would all see our safety degrade as the E911 system loses its enonomies of scale. This isn't something people think about, either, until you really need it.
So given that we have a technical solution, it really doesn't cost that much to implement, and it's in society's interest to ease access to emergency services, what's the problem?
Test users' appetite for downloads?? Why don't they just ask Apple or an Apple shareholder? Better yet, put the old Hitchhikers Guide or Dr. Who episodes online and they'll find out right quick about appetites for downloads!
Create a new account on your favorite free email server, like Yahoo! or Hotmail! or whoever just for the vacation and give that to people in advance. Tell people not to send sensitive stuff, but if something important comes up to have messages like "Please call Elsa. It's important." You can even leave a .vacation.msg that gives proper instructions on the use of te throw-away account.
This doesn't work for every use, but I could definitely understand the need to have some means of communication when one is on the go. This is especially true when going from country to country, say, in Europe, where I live.
In fact it was important for a friend who was bouncing between Germany and Switzerland who was having trouble connecting with a host. If you're going to be away so long that you need to do more extensive (and sensitive) stuff then I agree with the other posters that you should either bring a laptop or use smart technology. In fact, see my article about a proposed solution here.
On its own I wouldn't discount what Semantec says. However, "simulations" generally involve models, and those models have assumptions. What are the assumptions in this model, I wonder? We already know that a virus can travel roughly at the speed of a disk drive's ability to write.
It would be more interesting to see a study of computer-based virii versus biological ones. How about some real epidemiologists take a crack at it? Perhaps they already have..
Anyone? Anyone? Bueller?
While Londoners are using Microphones to handle noise violations, the people of East Palo Alto, California used them for accoustic analysis to determine from which direction the bullets were flying in a particularly gang-ridden neighborhood, and it was reported to have worked. Very few residents complained about their rights being violated.
Now, on the other hand, if London started recording conversations, that would make for a more interesting - and invasive- use. At that point we're heading right for 1984. Of course, London already uses cameras in public places for use as evidence, so their networking infrastructure is already there...
This report says absolutely nothing new. If you're going to take VoIP seriously, you need to recognize the application's needs. In this case, some amount of QoS is important, particularly at conjestion points such as the last hop to the consumer. You also need to recognize that like any other application on the Internet DDOS is a possibility. Ain't no different.
;-)
On the other hand, IPv6 will solve all our problems, right?
Microsoft breaking things doesn't help upgrades, indeed. Those sorts of security fixes are tricky, and sometimes the inherent functionality in use is the bug. Indeed the Mozilla folks just dealt with a similar matter for Firefox 1.0.3. For an amusing set of conversations check out Bugzilla bugs 289231 and 281988 where they successfully grappled with the same sorts of problems.
Terrestrially speaking, that formula's not quite right because no man made signal I know of goes through the center of the earth. You have to go around. The best you could possibly do at the equator is pi R in each direction or 2 pi R, which is > 6r.
That's about 0.133 ms.
At that bound VoIP is still good. In general codecs are good enough such that with buffering you can get away with up to 200ms without a human noticing. Beyond 200ms people start tripping over each other because of delay.
The gentleman who runs the Iraqi ISP mentioned that he could get 500ms round trip times. At best the people on the ends of these phones should demonstrate patience with the technology.
... whatever yore means, there was one service provider based in Texas whose management were quite offended by dirty pictures on USENET. So they hired a consultant to see just how much money that would save on... disk space, yes disk space.
The contractor came back and said that they could save 60% of their disk space, but since he also analyzed their NNTP logs he told them that porn was also their major source of revenue.
They're still around. Guess which path they chose.
Guys, it's nice and all that CNET, the WSG, and the Washington Post have all this info. Is there a link to the filing?
VoIP is still not a complete solution, at least not for reliable service just yet, IMHO
Drops are the result of networks that are not properly engineered for voice. You need to do that.
That means using appropriate QoS mechanisms at the edges and either using those same QoS mechanisms across the core or being sure to over-engineer the links (your mileage may vary).
VOIP is a tradeoff in this regard. Yes, you don't have to duplicate infrastructure, and you actually do get more efficient use of the bandwidth. In addition, the codecs used for IP phones are far more resilient to both loss and delay, as compared to an analog phone. An analog phone has no buffers to play with.
Heck, just post bits of your file system encrypted in a public key, and keep posting them. That way altavista can become your backup system.
Alright, a couple of things:
.org and yet come in well below the costs of other domains.
Not all of ISOC's programs benefit all of dot-org. So why tax all of dot-org?
Since the vast majority of ISOC's programs involve the IETF and standards organization support, it's really hard for me to accept the above statement.
The ones that do benefit all of dot-org benefit more than just dot-org. So why tax just dot-org?
I have no objection to taxing other domains, but I don't see the mechanism necessary to do the job. Furthermore, I'm actually betting that ISOC can "tax"
As for the rest of ISOC's stuff, find the right people to charge.
That's easy to say and hard to do. I would argue that since we all benefit from ISOC we should all pay a small fee, and so again I would accept the notion that all domain owners pay a small fee.
I would be happy to continue this conversation. To do so, disentangle my email address and feel free to mail me.
And I bet you also wouldn't want to pay to put out somebody elses' fire. But when it comes to YOUR house...
Want to get your document through the IETF process? Well, the IETF is going to need to scale. In order to do that, they're probably going to need money. And where will that money come from? How about those of us who benefit from their standards? Are we talking beaceaup bucks? Probably not, but I could certainly think of worse places for the money to go.
SSL/TLS is not appropriate if implemented atop TCP. While order is important, timeliness is more so.
If security of VOIP is no WORSE than that of a normal phone line, then at least you only have to worry about physically securing a single infrastructure as opposed to the multiple parallel ones you had to manage previously.
Also, in some environments VOIP on top of IPsec may be reasonable. The article is NOT entirely on target (IMHO it's a cheap hit piece). Consider the Cisco IP Phone in a work station. Since a person plugs his or her PC into the phone for network connectivity, you merely need to have some way to trust that the phone is authorized to use QoS, and you can thus encrypt voice traffic AND have the phone do classification.
And that trust can be gotten on the small with simple approaches such as MAC address lockdowns on your switches.
There was the incident of the fellow who discovered that the New York Times was left wide open by FrontPage. So he called to tell them, and was promptly arrested. I wonder if Mr. Clarke thinks that's fair.