Network addressing is hierarchical, so we will probably be running out of IPv6 addresses long before ~2^48 subnets are allocated. Of course that is still a few trillion.
You couldn't make a router big enough to process 2^48 subnets (let alone 2^64) directly. The hierarchy is necessary, and hierarchy means lots of "wasted" address space. To say nothing of the fact that the right most 2^64 bits of each IPv6 address aren't really routable at all - not without playing games at any rate.
So the whole idea that IPv6 supports 2^128 addresses in real life is some sort of cosmic joke. I bet it won't be a hundred years before people are chafing at the limitations of the IPv6 addressing system.
IP filtering has always been useless from a security standpoint. Same goes for MAC address filtering. Anyone anywhere can change both easily
Unless you have unfiltered BGP access to a major backbone, you have no hope of conducting a real conversation over the Internet using someone else's IP address, because the return packets will be routed back to them, not to you.
The ITU defines what 3G and 4G mean, and who can call their services by those names.
No, they don't. The ITU defines what "IMT-Advanced" is. They do not define what "4G" means, nor do they have the tiniest smidgen of legal authority to regulate who can use "4G" to refer to their services.
Realistically speaking, the ITU program even to define what IMT-Advanced is a complete waste of time. It is not an actual technical or interoperability standard, just some sort of bureaucratic bragging right. Butterfly classification at best. The ITU appears just trying to justify its existence, by pretending to be a regulatory authority.
It makes a bit of a difference whether you actually pay for a cloud based service or not. If something is free, you can't really complain when it goes away. Gift horse, mouth, etc.
If you don't like the rates being charged, you should purchase a generator, or solar panels, or a windmill, or...
New plan. The real value of the millions of miles of utility easements across and under public lands and streets is astronomical. So the government charges fair market value to all utilities. Fair market value is what they could get if they had to lease the easements from private land owners.
Then, as a matter of public policy, the government gives tax credits to those utilities with service and pricing policies that vaguely accord with the public interest in universal service at reasonable rates.
Either way the cable companies and the telcos are leeches at the public trough.
But the resource isn't actually scarce. I paid for unlimited usage of a 30-mbps link.
That is a bit of a non sequitur, don't you think? What ISPs lie and mislead people about has no bearing on whether they have the resources to deliver on their promises. Think Bernie Madoff.
The market rate for actually (not just stochastically) unlimited connections in the 1 Gb/s range is somewhere around $10 / mbps / month in the United States. Plus local loop charges.
The fact that you are not getting charged at least $300 / month for a 30 mbps "unlimited" connection simply means that you either do not run your connection anywhere near capacity or you are being wildly cross-subsidized by much lighter users.
And it won't be cheap; I would not rule out several dollars per gigabyte.
At rates that high, wireline providers would be courting FCC price regulation. I am not sure they could get away with charging a marginal rate of even $1 / GB at any reasonable level. What price the government will allow a monopoly or near monopoly to charge is highly dependent on the actual economics - in this case the actual economics of expanding the access networks to handle a much higher level of traffic. Plus a healthy margin on top of that. Public utilities regulation (or the threat of it), Internet style.
which is obviously not being done when you have a government agency telling companies what they can charge for services they provide.
Suppose the local electric company decides to charge $1 per kwh on weekdays, $2 per kwh on Saturdays, and $3 per kwh on Sundays and holidays. Should the local public utilities commission have the power to prevent that?
The local loop usually means you're using some monopoly telco and they can charge whatever they like.
That is what we have the FCC for - to get after the local access providers if they exploit their wireline monopolies too much. $0.10 or $0.20/GB marginal rate is not going to be a problem. If large wireline providers try to charge ten times that much (and you know they want to), that would probably bring Congress and the FCC bearing down on them with telecom style pricing regulation.
I agree. The other side of the story is distinguishing sick and vacation days is a form of health insurance that insures against the cost of being more sick than the average employee.
Of course, like most forms of insurance, it is subject to abuse, and tends to require auditing to make sure that people don't do just that. Far better to just treat employees fairly with a fixed amount of paid time off and let them make the call as to when and where.
[The comment form is not working very well in Chrome at the moment, btw]
Flash is the epitome of everything HTML5 is not, standards wise. Half the point of HTML5 (and related standards like SVG) is to make plugins like Flash and Silverlight unnecessary for open websites.
So if say a 10px letter would be the minimum readable size on the first iPhone, it's probably not readable at all on the latest iPhone model.
HTML/CSS renderers are supposed to use a virtual pixel size (nominally ~96 dpi on desktops) so that a major change (a doubling for example) in display device dpi doesn't cause a pixel specified page to render at a radically different size.
Without the litany of restrictions that come with Oracle/Sun's patent licensing, Java is more "fake open source" than real. The same goes for any implementation that doesn't come with the open licenses to the patents necessary to implement the specification in any field of use, domain of application, or open source alternative, compatible or not. Dalvik dispute, q.v.
I prefer that the government break up monopolies like Comcast, rather than set the prices used by monopolies to protect their monopoly and its abuses.
As long as video channels are transmitted over dedicated frequency bands on cable television systems, there is not a lot the government can do. Now that we are in a facilities based provider world (in large part due to some poor decisions by the FCC) the government has to regulate ISPs to a certain extent.
I would much prefer provider independent municipal fiber networks like UTOPIA, but the economics are currently relatively weak. I would pay the full cost of installing fiber to my residence, but my neighbors not so much.
What the FCC can, and should, prohibit is price gouging by networks with a monopoly on access between endpoints and other networks
I agree. I suspect FCC regulation of traffic settlement charges on a sender pays basis between major ISPs is likely, for the very reason you cite, and I think that is probably good thing.
The main thing I worry about FCC rate regulation in a case like this is government regulators tend to be too slow in reducing regulated rates to match falling economic costs. The beneficiaries of higher rates don't want them to.
Net Neutrality doesn't have anything to do with not charging for traffic. It has to do with arbitrarily blocking, filtering and prioritizing traffic based on type, source, and content. In the real world ISPs and telecom companies alike engage in traffic settlements. There is not a chance that the FCC is going to prohibit them, because (1) there is a decades old precedent from the telecom world, and (2) to do so would place the economics of the Internet at extraordinary risk. I am a tier 3 ISP and I would like free peering please. Not.
Care to support that with a legal argument? Because the entire FCC authority over the Internet hinges on whether ISPs are common carriers, covered by Title II of the Communications Act. The only way ISPs can avoid being classified as common carriers is for Congress to enact a statute so declaring.
But can someone explain to me why IPv6 didn't just extend the IPv4 format logically and stylistically? Why not just tack on more numbers? And all existing numbers could be assumed
The reason is that IPv4 wasn't designed to allow variable length addresses, so new IP version whatever hosts with VLA support still couldn't exchange traffic in both directions with IPv4 hosts, not without some sort of NAT in between. Small fixed length addresses are IPv4's fatal flaw.
They have already been paid by their customers for this.
That is a nice theory, but in practice peering arrangements between major providers don't work like that, but rather on a unpaid peering basis when traffic is roughly symmetric, or paid peering on a sender pays basis when traffic ratios are wildly asymmetric.
There are a number of technical reasons why that is the case, most notably the inability to identify who ultimately caused a packet to be transmitted on high capacity links. "Sender pays" is much simpler, and that is why it predominates. If the FCC gets involved they are not likely to decide differently.
Alternatively, due to Comcast's monopoly abuse, NetFlix and Akamai were absorbing costs that, in a fair market, would be absorbed by Comcast and the consumer.
Not true. The Internet is not a "caller pays" or "requester pays" system, because there is no way to make that differentiation on high capacity links for a variety of reasons. At high enough traffic volumes between major providers, it tends to be a "sender pays" system, sort of like the postal system.
Netflix generates an enormous amount of traffic, so they have to pay their ISP, which must then pay other major ISPs that they want to send wildly asymmetric levels of traffic to. Comcast is not filtering Netflix, nor discriminating against Netflix or Level 3, just maintaining standard Internet peering practices.
This is not a violation of net neutrality for that reason. If the FCC gets involved they are likely to ratify just the sort of traffic settlement practices that exist in the telco world, which are similar in principle and in practice to what goes on in peering agreements between major providers. In short, Level 3 is going to lose.
Yes, but ISPs also pay each other for bandwidth, and for large enough ISPs the rule is generally "sender pays", unless traffic volume in each direction is similar, where it is more common to have an unpaid peering arrangement.
Here Level 3 wants to dump an enormous amount of additional traffic on Comcast's network, so Comcast wants to be paid for the extra load. That of course causes in increase in Level 3's costs, which ultimately causes an increase in Netflix's costs. Either way the customer ultimately pays, either in increased Netflix service fees or in increased Comcast connection charges.
As long as Comcast is not filtering or blocking traffic, they appear to have the better of the argument here, and Level 3 / Netflix will almost certainly have to pay something resembling market rates to send very high volumes of traffic into Comcast's network.
Patents are neither about exotic, nor strictly about invention. It's mostly about innovation
Granting monopolies on innovations is an extraordinary impediment to further innovation. It doesn't matter what patents are "about" when their primary effect is to prevent progress in science and the useful arts, rather than to advance it. This is nowhere more obvious than in the field of software development.
Slashdot Mirror
IPv6 has 2^128 addresses
Network addressing is hierarchical, so we will probably be running out of IPv6 addresses long before ~2^48 subnets are allocated. Of course that is still a few trillion.
You couldn't make a router big enough to process 2^48 subnets (let alone 2^64) directly. The hierarchy is necessary, and hierarchy means lots of "wasted" address space. To say nothing of the fact that the right most 2^64 bits of each IPv6 address aren't really routable at all - not without playing games at any rate.
So the whole idea that IPv6 supports 2^128 addresses in real life is some sort of cosmic joke. I bet it won't be a hundred years before people are chafing at the limitations of the IPv6 addressing system.
IP filtering has always been useless from a security standpoint. Same goes for MAC address filtering. Anyone anywhere can change both easily
Unless you have unfiltered BGP access to a major backbone, you have no hope of conducting a real conversation over the Internet using someone else's IP address, because the return packets will be routed back to them, not to you.
The ITU defines what 3G and 4G mean, and who can call their services by those names.
No, they don't. The ITU defines what "IMT-Advanced" is. They do not define what "4G" means, nor do they have the tiniest smidgen of legal authority to regulate who can use "4G" to refer to their services.
Realistically speaking, the ITU program even to define what IMT-Advanced is a complete waste of time. It is not an actual technical or interoperability standard, just some sort of bureaucratic bragging right. Butterfly classification at best. The ITU appears just trying to justify its existence, by pretending to be a regulatory authority.
It makes a bit of a difference whether you actually pay for a cloud based service or not. If something is free, you can't really complain when it goes away. Gift horse, mouth, etc.
If you don't like the rates being charged, you should purchase a generator, or solar panels, or a windmill, or...
New plan. The real value of the millions of miles of utility easements across and under public lands and streets is astronomical. So the government charges fair market value to all utilities. Fair market value is what they could get if they had to lease the easements from private land owners.
Then, as a matter of public policy, the government gives tax credits to those utilities with service and pricing policies that vaguely accord with the public interest in universal service at reasonable rates.
Either way the cable companies and the telcos are leeches at the public trough.
But the resource isn't actually scarce. I paid for unlimited usage of a 30-mbps link.
That is a bit of a non sequitur, don't you think? What ISPs lie and mislead people about has no bearing on whether they have the resources to deliver on their promises. Think Bernie Madoff.
The market rate for actually (not just stochastically) unlimited connections in the 1 Gb/s range is somewhere around $10 / mbps / month in the United States. Plus local loop charges.
The fact that you are not getting charged at least $300 / month for a 30 mbps "unlimited" connection simply means that you either do not run your connection anywhere near capacity or you are being wildly cross-subsidized by much lighter users.
And it won't be cheap; I would not rule out several dollars per gigabyte.
At rates that high, wireline providers would be courting FCC price regulation. I am not sure they could get away with charging a marginal rate of even $1 / GB at any reasonable level. What price the government will allow a monopoly or near monopoly to charge is highly dependent on the actual economics - in this case the actual economics of expanding the access networks to handle a much higher level of traffic. Plus a healthy margin on top of that. Public utilities regulation (or the threat of it), Internet style.
If things get too bad, municipal shared access fiber is much more likely.
which is obviously not being done when you have a government agency telling companies what they can charge for services they provide.
Suppose the local electric company decides to charge $1 per kwh on weekdays, $2 per kwh on Saturdays, and $3 per kwh on Sundays and holidays. Should the local public utilities commission have the power to prevent that?
The local loop usually means you're using some monopoly telco and they can charge whatever they like.
That is what we have the FCC for - to get after the local access providers if they exploit their wireline monopolies too much. $0.10 or $0.20/GB marginal rate is not going to be a problem. If large wireline providers try to charge ten times that much (and you know they want to), that would probably bring Congress and the FCC bearing down on them with telecom style pricing regulation.
Change "sick" days to "personal" days.
I agree. The other side of the story is distinguishing sick and vacation days is a form of health insurance that insures against the cost of being more sick than the average employee.
Of course, like most forms of insurance, it is subject to abuse, and tends to require auditing to make sure that people don't do just that. Far better to just treat employees fairly with a fixed amount of paid time off and let them make the call as to when and where.
[The comment form is not working very well in Chrome at the moment, btw]
I think someone needs to gain an acquaintance with the Shannon Theorem.
Flash is the epitome of everything HTML5 is not, standards wise. Half the point of HTML5 (and related standards like SVG) is to make plugins like Flash and Silverlight unnecessary for open websites.
So if say a 10px letter would be the minimum readable size on the first iPhone, it's probably not readable at all on the latest iPhone model.
HTML/CSS renderers are supposed to use a virtual pixel size (nominally ~96 dpi on desktops) so that a major change (a doubling for example) in display device dpi doesn't cause a pixel specified page to render at a radically different size.
See here.
Er, that should be "With the litany of of restrictions..."
Java is Open Source
Without the litany of restrictions that come with Oracle/Sun's patent licensing, Java is more "fake open source" than real. The same goes for any implementation that doesn't come with the open licenses to the patents necessary to implement the specification in any field of use, domain of application, or open source alternative, compatible or not. Dalvik dispute, q.v.
I prefer that the government break up monopolies like Comcast, rather than set the prices used by monopolies to protect their monopoly and its abuses.
As long as video channels are transmitted over dedicated frequency bands on cable television systems, there is not a lot the government can do. Now that we are in a facilities based provider world (in large part due to some poor decisions by the FCC) the government has to regulate ISPs to a certain extent.
I would much prefer provider independent municipal fiber networks like UTOPIA, but the economics are currently relatively weak. I would pay the full cost of installing fiber to my residence, but my neighbors not so much.
What the FCC can, and should, prohibit is price gouging by networks with a monopoly on access between endpoints and other networks
I agree. I suspect FCC regulation of traffic settlement charges on a sender pays basis between major ISPs is likely, for the very reason you cite, and I think that is probably good thing.
The main thing I worry about FCC rate regulation in a case like this is government regulators tend to be too slow in reducing regulated rates to match falling economic costs. The beneficiaries of higher rates don't want them to.
Net Neutrality is not a toll road
Net Neutrality doesn't have anything to do with not charging for traffic. It has to do with arbitrarily blocking, filtering and prioritizing traffic based on type, source, and content. In the real world ISPs and telecom companies alike engage in traffic settlements. There is not a chance that the FCC is going to prohibit them, because (1) there is a decades old precedent from the telecom world, and (2) to do so would place the economics of the Internet at extraordinary risk. I am a tier 3 ISP and I would like free peering please. Not.
ISP's DO NOT HAVE COMMON CARRIER STATUS
Care to support that with a legal argument? Because the entire FCC authority over the Internet hinges on whether ISPs are common carriers, covered by Title II of the Communications Act. The only way ISPs can avoid being classified as common carriers is for Congress to enact a statute so declaring.
But can someone explain to me why IPv6 didn't just extend the IPv4 format logically and stylistically? Why not just tack on more numbers? And all existing numbers could be assumed
The reason is that IPv4 wasn't designed to allow variable length addresses, so new IP version whatever hosts with VLA support still couldn't exchange traffic in both directions with IPv4 hosts, not without some sort of NAT in between. Small fixed length addresses are IPv4's fatal flaw.
They have already been paid by their customers for this.
That is a nice theory, but in practice peering arrangements between major providers don't work like that, but rather on a unpaid peering basis when traffic is roughly symmetric, or paid peering on a sender pays basis when traffic ratios are wildly asymmetric.
There are a number of technical reasons why that is the case, most notably the inability to identify who ultimately caused a packet to be transmitted on high capacity links. "Sender pays" is much simpler, and that is why it predominates. If the FCC gets involved they are not likely to decide differently.
Alternatively, due to Comcast's monopoly abuse, NetFlix and Akamai were absorbing costs that, in a fair market, would be absorbed by Comcast and the consumer.
Not true. The Internet is not a "caller pays" or "requester pays" system, because there is no way to make that differentiation on high capacity links for a variety of reasons. At high enough traffic volumes between major providers, it tends to be a "sender pays" system, sort of like the postal system.
Netflix generates an enormous amount of traffic, so they have to pay their ISP, which must then pay other major ISPs that they want to send wildly asymmetric levels of traffic to. Comcast is not filtering Netflix, nor discriminating against Netflix or Level 3, just maintaining standard Internet peering practices.
This is not a violation of net neutrality for that reason. If the FCC gets involved they are likely to ratify just the sort of traffic settlement practices that exist in the telco world, which are similar in principle and in practice to what goes on in peering agreements between major providers. In short, Level 3 is going to lose.
Yes, but ISPs also pay each other for bandwidth, and for large enough ISPs the rule is generally "sender pays", unless traffic volume in each direction is similar, where it is more common to have an unpaid peering arrangement.
Here Level 3 wants to dump an enormous amount of additional traffic on Comcast's network, so Comcast wants to be paid for the extra load. That of course causes in increase in Level 3's costs, which ultimately causes an increase in Netflix's costs. Either way the customer ultimately pays, either in increased Netflix service fees or in increased Comcast connection charges.
As long as Comcast is not filtering or blocking traffic, they appear to have the better of the argument here, and Level 3 / Netflix will almost certainly have to pay something resembling market rates to send very high volumes of traffic into Comcast's network.
Patents are neither about exotic, nor strictly about invention. It's mostly about innovation
Granting monopolies on innovations is an extraordinary impediment to further innovation. It doesn't matter what patents are "about" when their primary effect is to prevent progress in science and the useful arts, rather than to advance it. This is nowhere more obvious than in the field of software development.