Agreed. NAT isn't a permanent solution. I disagree that sooner is better though. As with anything, the most cost effective transition will begin on its own when the time is right.
I don't know what you mean by buying infrastrcture. We're not losing out on any technology or experience really. If any important services become IPv6 only... well then we'd have a little catch-up--but that is precisely what will deliver the consumer demand.
CISCO is right in their problem prediction but they want to accelerate the timing so as to make money now, not later. Money now == more valuable.
I suggest you re-examine the history of electronic mail and then re-evaluate your understanding of what it means to be a network of networks...
It does not in fact merely mean routable ip networks. The internet was meant to bridge many networks that did not use IP by means of a gateway hosts that did speak IP.
I agree that no one specifically was thinking of NAT as we know it when network of networks was coined, but it is a simple extension of the principle.
There was a strong message in there that the problem with the current design is lack of identification of who is who. At least that's what I read into the business about phishing and spam.
The business about zombies seems like a potential code for the need to block "normal" users from connecting with each other.
Might be because we realized that the IPV6 protocol was unnecessary.
Once people were forced to NAT, it suddently dawned on the great mass of people that workstations shouldn't be getting public IPs for security and management reasons.
Nor for that matter should these up and coming embedded devices be placed on the public internet either. It just isn't appropriate.
Remember: The Internet was supposed to be a network of networks NOT _THE NETWORK_.
Most of the remaining IP allocation problems result from certain lingering gross misallocations such as the Class A block assigned to MIT.
No, its the "conservatives" who are most consistent liberal voices on the court. The conservatives (Thomas, O'Connor, et al) are about conserving traditional british liberalism---whereas the "liberals" are about conserving statism.
You're buying into a canard if you think that the conservative position is "states rights". The conservative position is more often "explcit federal rights, defer to the states when heighened ambiguity exists."
So... your definitions of conservative and liberal are all screwy. By this day and age, statists ARE conservatives too.
If you escape the bubble of contemporary US politics you begin to understand that
1) liberal and conservative are not opposites 2) that a conservative should describe anyone who is defending a tradition or is at least initially biased against change. 3) that a liberal should describe anyone who supports individual rights 4) that a libertarian is a very particular kind of liberal 5) that leftist is not necessarily liberal but might be very statist
Then: 6) that Breyers, Ginsburg, et al often vote the statist position and only on a handful of sweeteners do they do vote to protect individual rights 7) that Scalia, Thomas, Connor, and Rehenquist often vote in favor of broad liberties and when they do not, it is not because they are opposing the liberties but because they oppose the Federal government dictating too many details.
Way to rag on hardware designers. I think we'll take our computers back now. Go back to your mechanical adding machines.
Seriously: have you see the code "programmers" make? Hardware designers tend to be extremely effective programmers--if only because they have a clue as the actual performance costs of their design decisions. Most programmers haven't got a clue that an IPI stalls all the activity on an SMP system let alone a reasonable sense of how costly that is.
From what I've seen, computer engineers get as much training is software engineering methodology as programmers.
The *trouble* is tight schedules and billion things you have to do.
Actually that was surprising arrogant of him. It suggests that he feels he has nothing to learn by studying alternative designs wrt to their realized performance.
There was certainly a time when Linux did have a lot to learn from FreeBSD (but I would say that 2.6 essentially ended that).
For example, it is well known that Matt Dillon (then of FreeBSD) advised the linux kernel team on how to fix their VM, and further that FreeBSD's Alan Cox helped dig them out of the VM mess they made when attempting to implement those ideas.
The subtle subtext of Linus's commentary is his belief in a monoculture. "One kernel to rule them all." This very much mirrors the distinction in focus that I perceive between the linux community and the BSD community in general. The latter is focused much much more on exploring different Kernel architecture decisions whereas the former is much much more focused on exploring different userland structures.
There are some really radical conceptual differences between the Kernels in NetBSD (BGL), FreeBSD 5 (Fine-grained mutexes), Dragonfly (Serializing Tokens, cpu-localization), FreeBSD 4 (highly-optimized uniprocessor performance w/ BGL smp)
Actually the answer is that the manufurerer, the consumer, and the reseller all pay in proportion to their price elasticities.
The secret is that the government is not able to actually direct the tax--unless the government regulates the prices and behavior of everyone involved. It can only defraud the public by counterfactually suggesting that the tax is directed.
Actually IBM has a C and Fortran compiler they made just for that reason--whether it really works well, that's hard to tell at this point.
And I think you grossly misunderstand what's wrong with the Itanium. It had *nothing* to do per se with permitting the compiler to organize the fine-grained parallelism. It had to do with throwing in the kitchen sink so to speak of every clever idea people have had--big surprise they don't all work out well together.
But the major problem was placing so much additional load on the memory interface by using such a large register set and using such large instructions.
Nono. They are just fine for algorithmic or complex flow--as good as a powerpc is, but this is a poor utilization of the hardware.
The real limitation of the Cell processor would be context switches. There is a hugh register file to take-care of in that case.
Sadly, this processor will only be useful in graphics applications. It's double-precision floating-point performance isn't very good (about 1/10 of the speed)--all us science types really value double precision performance...
as for accessing data in awkward ways... it isn't so bad... it's more like you have to manually drive the caching strategy. Some might see this as a good idea when the normal sequential heuristic used in caches isn't good enough.
Anyways, I was at this demo. The graphics themselves weren't too impressive but the frame-rate on the ray-tracing application was very good--about 10x faster than their version for the normal powerpc.
You missed the most important of all: mothball the international space station. The ISS is killing NASA... and it's not just the space-station itself. The ISS is the prime motivation driving the shuttles back into service.
This is still a very usual split, but overall, the political nature of the court is exaggerated. When I last saw stats (several years ago), any pairwise comparison of the justices found them in agreement at least 60% of the time.
That said, this particular 5-4 split has not happened in the past ten years
However, within 5-4 decisions, Stevens and Thomas agree about 16 percent of the time.
Scalia was clearly the swing-vote on this case
"Although Scalia is no fan of the dormant commerce clause, he has written that:... I will, on stare decisis grounds, enforce a self executing "negative" Commerce Clause in two situations: (1) against a state law that facially discriminates against interstate commerce, and (2) against a state law that is indistinguishable from a type of law previously held unconstitutional by this Court.
Since the state laws in question here demonstrably fell into the former category, and we can infer that Scalia was not persuaded by Thomas' account of the 21st amendment, stare decisis required him to vote to strike down these laws." (http://www.professorbainbridge.com/2005/05/suprem e_court_s.html)
Be fair. Most graduate students in CS or EE are paid for their research (although not very much) and given free tuition.
I agree most of what people publish and talk about is crap. A conference of 50 papers is likely to have only a couple ones that are probably on track. Last year, a guy was presenting on a new asynchronous logic family he had invented. He compared it to a family I had introduced three years earlier--i.e., was bragging that his was faster. His scheme of course was _my scheme_ minus all the details I had included to tolerate process parameter variations. So of course his went slightly faster, he squeezed out the engineering margin. Mind you that I had remarked in my original paper... "if we remove these two transistors, the circuit will operate slightly faster but be much more dependent on exactly knowing the process and operating characteristics." And guess what he did?
Luckily my talk was scheduled after his. I threw away most of what I had prepared and instead lectured about his dangerous misunderstanding. Immunity to process parameter variation is essential--no one from industry is going to buy into something that isn't reliable at all four process corners of the P/N drive strength variation.
Let me tell you, peer review is _broken_. Reviewers almost never both to look into the references given by the paper or perform any sort of literature search against the paper topic.
Let me fill in some holes others left out. New releases are announced as having fixed security problems, but that is a comparison against the previous version's original ISO image only.
Security fixes are backported to earlier versions. Those versions still officially maintained have fixes backported by the security officers. Older versions tend to also get fixes but merely by the work of interested committers. Thus it isn't usual to see fixes being backported to releases as far back as 4.3.
What do I mean by backported? Users can update their/usr/src directory and rebuild. More recently a binary update service has been available.
Thus there is for example 5.3-RELEASE, and 5.3-p5.
Generally speaking, there is no need to wait for new releases to get fixes. Fixes are painlessly and automatically available almost overnight.
All of this applies to the software officially maintained by the FreeBSD system--i.e., anything in the "base system" Other software generally gets fixes in ports soon after the upstream version has a fix... but backing this is the port-audit database. port-audit is maintained by the security team and lists all the known vulnerabilities against third-party software. A cron job mails you warnings about vulnerable third-party software. The ports system warns you about vulnerable software and libraries when you attempt to install (even when a new install depends on an already installed but vulernable library.
> bigger and badder kill messages Quoting from the trade press huh? I usually find that it is pretty easy to know whether a service has shutdown.
If you look into the rcng architecture on FreeBSD, Dragonfly, and others, you'll see that the rc system provides a "start, stop, restart" command interface for most services. More over, it can start programs in parallel and auto-resolve dependencies.
All of this is provided in a very extensible way (it's written in sh). I wouldn't call this a specific unix trick or gotcha.
RCNG leverages the _power_ of the unix way (tm) to make an extensible interface.
Um. I am not opposing the 911 system. Your whole line of commentary about the "dispatch center" is a complete red-herring misinterpretation of my comments.
Remember, 911 usually maps to local police emergency numbers (or vice-versa depending on your point of view). The question is why is the number 911. I explained why. I for one remember getting by just fine before 911 existed.
Anyways, my concern in mentioning routing issues is a doubt in the ability of a random 911 centers to reroute a call about an issue in an arbitrary location. I was merely rebuting the suggestion that the call could be routed anywhere and things would work out.
I agree the center *still* doesn't know your position. I even mentioned that, but my reaction is so? What's your point? If you value the 911 center being able to know where you are without you're telling them, then get a land-line.
Anyways, I was trying to offer what I considered to be a low-cost+feasible answer to offering 911 like services on VoIP. I don't think GPS is feasible or reasonable.
To summarize, a little extra dialing can give many of the benefits without too much extra cost, though it doesn't have all the benefits.
It's a compromise. Thus the "feasible" part. It offers some of the benefits (versus the problem 911 is attempting to fix) and solves some of the implementation problems of just having 911.
It isn't clear that you would need to get precisely the area code for where you were. Getting an area code in the same state would bring you to the level of service of cellphone 911.
Fair enough. I didn't mean anything personal by replying to your comment in particular. It was more that your comment typified a particular proposal to the problem--thus it was a good point to pursue.
Okay. I happen to feel it isn't too burdensome to know the area code, but that might because the back of my day-calendar has an area code map of the united states.
As another form of this problem though, imagine you know that a crime is taking place some where you aren't. Wouldn't it be nice to be able to easily dial for a dispatcher in that area?
Alternatively... maybe you shouldn't use your VoIP phone to make that call?
Does every piece of technology need to solve every problem?
Last time I checked the appeal of VoIP was that it offered a cheap, simple way to speak to people...
You mean learning 6 or 7 digits instead of 3? More importantly 6 or 7 seven digits that follow a pattern. 1-xxx-911?
I agree we hear those stores occasionally... though confirmation bias on the part of the media ensures we hear about them disproportionately.
I doubt that the cost benefit analysis would go your way, but that's just my opinion...
And don't commit a generalization fallacy. I agree there are times when the simplist solution is ideal and nice, but that isn't a conclusive arguement for imposing a requirement that a particular implementation must be available.
So what exactly are you proposing? That the FCC require VoIP providers support transmission of location information? That the FCC require VoIP providers to properly route you to the local 911 center? That the FCC require the VoIP providers mandate that all VoIP services be location enabled?
What about people who don't want to stick a GPS dongle into their computer? What about people who *forget* to stick in the dongle when they rush to call 911? What about the people who claim to have done those things but haven't actually, and then turn around and sue?
Here's the trouble: It isn't that 911 operator needs to know where you are; you can give that information just fine. The trouble is that your call has to go to a local 911 dispatch center, not a dispatch center 3000 miles away.
If you've ever dealt with 911, then you know that they could never handle rerouting calls. Often it seems that they are barely able to properly dispatch local officers and emergency medical services.
Why do we have 911? People were observed to be stupid. The local police used to just have an emergency number. Doh, people can't remember the number. what to do, what to do... make one number for everyone, everywhere. make a big marketing campaign. yeah.
This whole problem would go away if you just had to give at least an area code as a prefix to dialing 911. Then the call could be routed to a local-state dispatcher. *Now that sounds feasible*.
As for your being beaten to death and you just barely manage to dial 911 and the police only hear the crime and trace the call... well that's sort of a fringe benefit of getting a land-land. Perhaps you should consider that before opting for other solutions.
Do we really all need to be burdened? Can some people just be free to offer bare-bones service?
Do you really want GPS tracking of your location--mandated by the government?
Slashdot Mirror
Agreed. NAT isn't a permanent solution. I disagree that sooner is better though. As with anything, the most cost effective transition will begin on its own when the time is right.
I don't know what you mean by buying infrastrcture. We're not losing out on any technology or experience really. If any important services become IPv6 only... well then we'd have a little catch-up--but that is precisely what will deliver the consumer demand.
CISCO is right in their problem prediction but they want to accelerate the timing so as to make money now, not later. Money now == more valuable.
I suggest you re-examine the history of electronic mail and then re-evaluate your understanding of what it means to be a network of networks...
It does not in fact merely mean routable ip networks. The internet was meant to bridge many networks that did not use IP by means of a gateway hosts that did speak IP.
I agree that no one specifically was thinking of NAT as we know it when network of networks was coined, but it is a simple extension of the principle.
You noticed that too?
There was a strong message in there that the problem with the current design is lack of identification of who is who. At least that's what I read into the business about phishing and spam.
The business about zombies seems like a potential code for the need to block "normal" users from connecting with each other.
Might be because we realized that the IPV6 protocol was unnecessary.
Once people were forced to NAT, it suddently dawned on the great mass of people that workstations shouldn't be getting public IPs for security and management reasons.
Nor for that matter should these up and coming embedded devices be placed on the public internet either. It just isn't appropriate.
Remember: The Internet was supposed to be a network of networks NOT _THE NETWORK_.
Most of the remaining IP allocation problems result from certain lingering gross misallocations such as the Class A block assigned to MIT.
No offense, but maybe you missed the point?
The article was entitled "Keeping a Data Center Cool _on the Cheap_" (emphasis added)
The suggestion being that they were able to significantly cut back on the size and power consumption of the cooling plant by using plastic wrap.
No, its the "conservatives" who are most consistent liberal voices on the court. The conservatives (Thomas, O'Connor, et al) are about conserving traditional british liberalism---whereas the "liberals" are about conserving statism.
You're buying into a canard if you think that the conservative position is "states rights". The conservative position is more often "explcit federal rights, defer to the states when heighened ambiguity exists."
So... your definitions of conservative and liberal are all screwy. By this day and age, statists ARE conservatives too.
If you escape the bubble of contemporary US politics you begin to understand that
1) liberal and conservative are not opposites
2) that a conservative should describe anyone who is defending a tradition or is at least initially biased against change.
3) that a liberal should describe anyone who supports individual rights
4) that a libertarian is a very particular kind of liberal
5) that leftist is not necessarily liberal but might be very statist
Then:
6) that Breyers, Ginsburg, et al often vote the statist position and only on a handful of sweeteners do they do vote to protect individual rights
7) that Scalia, Thomas, Connor, and Rehenquist often vote in favor of broad liberties and when they do not, it is not because they are opposing the liberties but because they oppose the Federal government dictating too many details.
Way to rag on hardware designers. I think we'll take our computers back now. Go back to your mechanical adding machines.
Seriously: have you see the code "programmers" make? Hardware designers tend to be extremely effective programmers--if only because they have a clue as the actual performance costs of their design decisions. Most programmers haven't got a clue that an IPI stalls all the activity on an SMP system let alone a reasonable sense of how costly that is.
From what I've seen, computer engineers get as much training is software engineering methodology as programmers.
The *trouble* is tight schedules and billion things you have to do.
Actually that was surprising arrogant of him. It suggests that he feels he has nothing to learn by studying alternative designs wrt to their realized performance.
0 41 9.html)
There was certainly a time when Linux did have a lot to learn from FreeBSD (but I would say that 2.6 essentially ended that).
For example, it is well known that Matt Dillon (then of FreeBSD) advised the linux kernel team on how to fix their VM, and further that FreeBSD's Alan Cox helped dig them out of the VM mess they made when attempting to implement those ideas.
(http://mail.nl.linux.org/linux-mm/2000-05/msg0
The subtle subtext of Linus's commentary is his belief in a monoculture. "One kernel to rule them all." This very much mirrors the distinction in focus that I perceive between the linux community and the BSD community in general. The latter is focused much much more on exploring different Kernel architecture decisions whereas the former is much much more focused on exploring different userland structures.
There are some really radical conceptual differences between the Kernels in NetBSD (BGL), FreeBSD 5 (Fine-grained mutexes), Dragonfly (Serializing Tokens, cpu-localization), FreeBSD 4 (highly-optimized uniprocessor performance w/ BGL smp)
Actually the answer is that the manufurerer, the consumer, and the reseller all pay in proportion to their price elasticities.
The secret is that the government is not able to actually direct the tax--unless the government regulates the prices and behavior of everyone involved. It can only defraud the public by counterfactually suggesting that the tax is directed.
Actually IBM has a C and Fortran compiler they made just for that reason--whether it really works well, that's hard to tell at this point.
And I think you grossly misunderstand what's wrong with the Itanium. It had *nothing* to do per se with permitting the compiler to organize the fine-grained parallelism. It had to do with throwing in the kitchen sink so to speak of every clever idea people have had--big surprise they don't all work out well together.
But the major problem was placing so much additional load on the memory interface by using such a large register set and using such large instructions.
Nono. They are just fine for algorithmic or complex flow--as good as a powerpc is, but this is a poor utilization of the hardware.
The real limitation of the Cell processor would be context switches. There is a hugh register file to take-care of in that case.
Sadly, this processor will only be useful in graphics applications. It's double-precision floating-point performance isn't very good (about 1/10 of the speed)--all us science types really value double precision performance...
as for accessing data in awkward ways... it isn't so bad... it's more like you have to manually drive the caching strategy. Some might see this as a good idea when the normal sequential heuristic used in caches isn't good enough.
Anyways, I was at this demo. The graphics themselves weren't too impressive but the frame-rate on the ray-tracing application was very good--about 10x faster than their version for the normal powerpc.
You missed the most important of all: mothball the international space station. The ISS is killing NASA... and it's not just the space-station itself. The ISS is the prime motivation driving the shuttles back into service.
This is still a very usual split, but overall, the political nature of the court is exaggerated. When I last saw stats (several years ago), any pairwise comparison of the justices found them in agreement at least 60% of the time.
... I will, on stare decisis grounds, enforce a self executing "negative" Commerce Clause in two situations: (1) against a state law that facially discriminates against interstate commerce, and (2) against a state law that is indistinguishable from a type of law previously held unconstitutional by this Court.
m e_court_s.html)
That said, this particular 5-4 split has not happened in the past ten years
However, within 5-4 decisions, Stevens and Thomas agree about 16 percent of the time.
Scalia was clearly the swing-vote on this case
"Although Scalia is no fan of the dormant commerce clause, he has written that:
Since the state laws in question here demonstrably fell into the former category, and we can infer that Scalia was not persuaded by Thomas' account of the 21st amendment, stare decisis required him to vote to strike down these laws." (http://www.professorbainbridge.com/2005/05/supre
Actually the result of this ruling is that states must ban all--regardless of origin--direct to consumer sales if they block them at all.
e _court_s.html
The supreme court merely ruled that states could not treat intra-state state sales differently from out of state sales.
The ruling preserves state control over this issue as long as the policy doesn't discriminate against out of state sellers.
see: http://www.professorbainbridge.com/2005/05/suprem
Be fair. Most graduate students in CS or EE are paid for their research (although not very much) and given free tuition.
I agree most of what people publish and talk about is crap. A conference of 50 papers is likely to have only a couple ones that are probably on track. Last year, a guy was presenting on a new asynchronous logic family he had invented. He compared it to a family I had introduced three years earlier--i.e., was bragging that his was faster. His scheme of course was _my scheme_ minus all the details I had included to tolerate process parameter variations. So of course his went slightly faster, he squeezed out the engineering margin. Mind you that I had remarked in my original paper... "if we remove these two transistors, the circuit will operate slightly faster but be much more dependent on exactly knowing the process and operating characteristics." And guess what he did?
Luckily my talk was scheduled after his. I threw away most of what I had prepared and instead lectured about his dangerous misunderstanding. Immunity to process parameter variation is essential--no one from industry is going to buy into something that isn't reliable at all four process corners of the P/N drive strength variation.
Let me tell you, peer review is _broken_. Reviewers almost never both to look into the references given by the paper or perform any sort of literature search against the paper topic.
Let me fill in some holes others left out. New releases are announced as having fixed security problems, but that is a comparison against the previous version's original ISO image only.
/usr/src directory and rebuild. More recently a binary update service has been available.
Security fixes are backported to earlier versions. Those versions still officially maintained have fixes backported by the security officers. Older versions tend to also get fixes but merely by the work of interested committers. Thus it isn't usual to see fixes being backported to releases as far back as 4.3.
What do I mean by backported? Users can update their
Thus there is for example 5.3-RELEASE, and 5.3-p5.
Generally speaking, there is no need to wait for new releases to get fixes. Fixes are painlessly and automatically available almost overnight.
All of this applies to the software officially maintained by the FreeBSD system--i.e., anything in the "base system" Other software generally gets fixes in ports soon after the upstream version has a fix... but backing this is the port-audit database. port-audit is maintained by the security team and lists all the known vulnerabilities against third-party software. A cron job mails you warnings about vulnerable third-party software. The ports system warns you about vulnerable software and libraries when you attempt to install (even when a new install depends on an already installed but vulernable library.
> bigger and badder kill messages
Quoting from the trade press huh? I usually find that it is pretty easy to know whether a service has shutdown.
If you look into the rcng architecture on FreeBSD, Dragonfly, and others, you'll see that the rc system provides a "start, stop, restart" command interface for most services. More over, it can start programs in parallel and auto-resolve dependencies.
All of this is provided in a very extensible way (it's written in sh). I wouldn't call this a specific unix trick or gotcha.
RCNG leverages the _power_ of the unix way (tm) to make an extensible interface.
Because Fortran has well-defined side-effect rules that easily permit loops to be unrolled into vector operations by a compiler.
(that's at least one reason).
Um. I am not opposing the 911 system. Your whole line of commentary about the "dispatch center" is a complete red-herring misinterpretation of my comments.
Remember, 911 usually maps to local police emergency numbers (or vice-versa depending on your point of view). The question is why is the number 911. I explained why. I for one remember getting by just fine before 911 existed.
Anyways, my concern in mentioning routing issues is a doubt in the ability of a random 911 centers to reroute a call about an issue in an arbitrary location. I was merely rebuting the suggestion that the call could be routed anywhere and things would work out.
I agree the center *still* doesn't know your position. I even mentioned that, but my reaction is so? What's your point? If you value the 911 center being able to know where you are without you're telling them, then get a land-line.
Anyways, I was trying to offer what I considered to be a low-cost+feasible answer to offering 911 like services on VoIP. I don't think GPS is feasible or reasonable.
To summarize, a little extra dialing can give many of the benefits without too much extra cost, though it doesn't have all the benefits.
Hardly. Which is harder to remember:
xxx-xxxx vs. 1xxx911 vs 911?
It's a compromise. Thus the "feasible" part. It offers some of the benefits (versus the problem 911 is attempting to fix) and solves some of the implementation problems of just having 911.
It isn't clear that you would need to get precisely the area code for where you were. Getting an area code in the same state would bring you to the level of service of cellphone 911.
Fair enough. I didn't mean anything personal by replying to your comment in particular. It was more that your comment typified a particular proposal to the problem--thus it was a good point to pursue.
Okay. I happen to feel it isn't too burdensome to know the area code, but that might because the back of my day-calendar has an area code map of the united states.
As another form of this problem though, imagine you know that a crime is taking place some where you aren't. Wouldn't it be nice to be able to easily dial for a dispatcher in that area?
Alternatively... maybe you shouldn't use your VoIP phone to make that call?
Does every piece of technology need to solve every problem?
Last time I checked the appeal of VoIP was that it offered a cheap, simple way to speak to people...
You mean learning 6 or 7 digits instead of 3? More importantly 6 or 7 seven digits that follow a pattern. 1-xxx-911?
I agree we hear those stores occasionally... though confirmation bias on the part of the media ensures we hear about them disproportionately.
I doubt that the cost benefit analysis would go your way, but that's just my opinion...
Trama is one thing. But preparedness is another. Knowing what the numbers are ahead of time. Programming them into your speed dial. Etc.
Anyways stupid is a broader word than you give it credit: "4. Dazed, stunned, or stupefied." (http://dictionary.reference.com/search?q=stupid)
And don't commit a generalization fallacy. I agree there are times when the simplist solution is ideal and nice, but that isn't a conclusive arguement for imposing a requirement that a particular implementation must be available.
So what exactly are you proposing? That the FCC require VoIP providers support transmission of location information? That the FCC require VoIP providers to properly route you to the local 911 center? That the FCC require the VoIP providers mandate that all VoIP services be location enabled?
What about people who don't want to stick a GPS dongle into their computer? What about people who *forget* to stick in the dongle when they rush to call 911? What about the people who claim to have done those things but haven't actually, and then turn around and sue?
Here's the trouble: It isn't that 911 operator needs to know where you are; you can give that information just fine. The trouble is that your call has to go to a local 911 dispatch center, not a dispatch center 3000 miles away.
If you've ever dealt with 911, then you know that they could never handle rerouting calls. Often it seems that they are barely able to properly dispatch local officers and emergency medical services.
Why do we have 911? People were observed to be stupid. The local police used to just have an emergency number. Doh, people can't remember the number. what to do, what to do... make one number for everyone, everywhere. make a big marketing campaign. yeah.
This whole problem would go away if you just had to give at least an area code as a prefix to dialing 911. Then the call could be routed to a local-state dispatcher. *Now that sounds feasible*.
As for your being beaten to death and you just barely manage to dial 911 and the police only hear the crime and trace the call... well that's sort of a fringe benefit of getting a land-land. Perhaps you should consider that before opting for other solutions.
Do we really all need to be burdened? Can some people just be free to offer bare-bones service?
Do you really want GPS tracking of your location--mandated by the government?