You could have extended the address space without doing "ships in the night". Add an optional field with "original source IP address" which NAT-routers fill in when they translate addresses. The optional source route fields would be good candidates, since they're universally ignored these days. Hosts answering traffic with "original source IP" set, reflect the value into the "original destination IP" field. When the NAT receives traffic from outside, they check if the "original destionation IP" field is present, and if so, they don't use their normal NAT state tracking but instead just replaces the destination address with whatever was in the "original destionation IP" field.
This would mean updates to all hosts and all NAT devices, but connectivity wouldn't break until this is in place, it would just revert to traditional NAT. Non-NAT-routers wouldn't need any updates.
Apple Airport. With all the disclaimers about being an Apple product and not being able to configure it through HTTP or SSH. Still, Apple seems to be the only consumer CPE vendor who is actively participating in the development of the various IPv6 CPE RFC's. (How is that for acronym soup?)
Once the IPv4 space runs out, it is likely that there will be a secondary market for/24's. That/8 could make quite a bit of money if carved into/24's. It would also be another 65k routes for the global routing table, which would be no fun at all.
We should appreciate their gesture; they did lose something and we are all slightly richer because of what they did.
At some point the powers-that-be will simply say "you're not using it, give it back"
Which powers would that be? The various registries can appeal to the community, but when it comes down to it, every ISP picks which routes they accept.
If you want to force the hands of the ISPs, you need legislation. World-wide, if you want it to be effective.
Except those of us forced to run Oracle Java because Java programmers are universally clueless and so the online banking applet doesn't work in OpenJDK. First you need to build the RPM yourself because JRE 1.6.0 STILL isn't free software (or accept the prebuilt pseudo-RPM which craps all over the system), and after that there are no more automatic updates.
Just to REALLY make sure that you might not accidentally lose out at the chance of some malware, Oracle Java also needs stack execution protection turned off in Firefox.
Danish studies have shown clear correlation between how well informed people are about GMO and how positive their attitude is towards it. The correlation is negative.
Polarized glasses mean that we have to put a filter in front of the TV which is able to switch polarization for each frame. I haven't heard of such technology.
Alternatively you'd need to halve resolution and put a fixed filter in front of each pixel. Actually, that seems like a really good idea now that I think about it. You might need OLED screens to be able to pull off the required resolution (prefererably 3840x1080 for Full-HD), but it would eliminate several of the disadvantages of current technology.
Apple Cinema Display. Dell has a version too. Both have been on the market for years, in various incarnations. And don't complain about the price, any CRT which can actually go black->white horizontally at 2500 pixels isn't cheap and never was.
3D is practically free for a modern TV, as long as the manufacturer doesn't care about the awful bleed-through between the frames. All they need to do is allow a reasonably high refresh rate, fiddle a little with the HDMI input, and send out the VSYNC signal somewhere so the glasses can access it.
The glasses on the other hand are expensive, so it makes sense that no one buys them.
At least this Google TV will use little enough power that you won't feel completely stupid for having it turned on in 3 years, and it doesn't cost double what a comparable TV without Android would cost. In fact, the additional cost seems to be very close to 0.
What you have proposed has been done several times, typically with Windows Media Center. They didn't sell, for obvious reasons.
Software (or rather, the OpenGL or Direct3D driver) has absolutely no problem rendering everything 90 degrees transformed. 2D is a different problem, but 2D is trivial in software on any remotely recent system. You can also render it on a texture and let the 3D hardware handle the problem.
Ok I don't get this one. Are you saying that I get worse frame rates when I rotate 90 degrees so the horizon is vertical in a flight simulator, with the cockpit turned off? Or does it take longer to paint the static 2D cockpit view the other way around?
Similarly with lying on your side in an FPS, but most of them probably don't allow that.
I'm not sure what you're on, but reading narrow columns is way faster than reading wide lines. That's why newspapers have columns. One of the many deficiencies of CSS is that it's practically impossible to a newspaper-like layout which works at any screen size (adapting the number of columns as needed).
Where did this obsession with Widescreen come from anyways? I understand for "widescreen films", but why are all monitors wide now? It's weird that it kind of slowly crept into the norm..
It's dead convenient to have one wide monitor rather than two narrow ones. Screens still aren't quite as wide as I could dream of, but we're getting there. 2:1 would be a good start.
Don't worry, the insecurity of current OS's and the ubiquity of NAT in IPv4 has forced IPv6 CPE's to block all incoming traffic. This is not quite as bad as NAT, but on the other hand it doesn't support UPNP. Good luck getting SIP to work across that.
I was in the middle of a long explanation about how your 64k is wrong, but there are much more fundamental problems with your understanding. For a starter, your understanding of RFC1918 is flawed.
Slashdot Mirror
Good luck NAT-ing four billion IP addresses behind one NAT box which has one IP address and 65536 ports.
This is only a problem if you're Cisco. Sane NAT implementations track connections by the whole 5-tuple.
You could have extended the address space without doing "ships in the night". Add an optional field with "original source IP address" which NAT-routers fill in when they translate addresses. The optional source route fields would be good candidates, since they're universally ignored these days. Hosts answering traffic with "original source IP" set, reflect the value into the "original destination IP" field. When the NAT receives traffic from outside, they check if the "original destionation IP" field is present, and if so, they don't use their normal NAT state tracking but instead just replaces the destination address with whatever was in the "original destionation IP" field.
This would mean updates to all hosts and all NAT devices, but connectivity wouldn't break until this is in place, it would just revert to traditional NAT. Non-NAT-routers wouldn't need any updates.
Apple Airport. With all the disclaimers about being an Apple product and not being able to configure it through HTTP or SSH. Still, Apple seems to be the only consumer CPE vendor who is actively participating in the development of the various IPv6 CPE RFC's. (How is that for acronym soup?)
Once the IPv4 space runs out, it is likely that there will be a secondary market for /24's. That /8 could make quite a bit of money if carved into /24's. It would also be another 65k routes for the global routing table, which would be no fun at all.
We should appreciate their gesture; they did lose something and we are all slightly richer because of what they did.
At some point the powers-that-be will simply say "you're not using it, give it back"
Which powers would that be? The various registries can appeal to the community, but when it comes down to it, every ISP picks which routes they accept.
If you want to force the hands of the ISPs, you need legislation. World-wide, if you want it to be effective.
The Linux version listens to DBUS events, so it knows whether the system thinks it is online or not.
Except those of us forced to run Oracle Java because Java programmers are universally clueless and so the online banking applet doesn't work in OpenJDK. First you need to build the RPM yourself because JRE 1.6.0 STILL isn't free software (or accept the prebuilt pseudo-RPM which craps all over the system), and after that there are no more automatic updates.
Just to REALLY make sure that you might not accidentally lose out at the chance of some malware, Oracle Java also needs stack execution protection turned off in Firefox.
Niche markets like "Europe". It's only now with semi-automatic gearboxes that non-manuals are becoming slightly more common.
Danish studies have shown clear correlation between how well informed people are about GMO and how positive their attitude is towards it. The correlation is negative.
Polarized glasses mean that we have to put a filter in front of the TV which is able to switch polarization for each frame. I haven't heard of such technology.
Alternatively you'd need to halve resolution and put a fixed filter in front of each pixel. Actually, that seems like a really good idea now that I think about it. You might need OLED screens to be able to pull off the required resolution (prefererably 3840x1080 for Full-HD), but it would eliminate several of the disadvantages of current technology.
Apple Cinema Display. Dell has a version too. Both have been on the market for years, in various incarnations. And don't complain about the price, any CRT which can actually go black->white horizontally at 2500 pixels isn't cheap and never was.
The standard is just double frame rate video, it's difficult to imagine that it could change significantly.
3D is practically free for a modern TV, as long as the manufacturer doesn't care about the awful bleed-through between the frames. All they need to do is allow a reasonably high refresh rate, fiddle a little with the HDMI input, and send out the VSYNC signal somewhere so the glasses can access it.
The glasses on the other hand are expensive, so it makes sense that no one buys them.
This HAS been used successfully in a file sharing defence.
You want to replace your TV every 3 years?
At least this Google TV will use little enough power that you won't feel completely stupid for having it turned on in 3 years, and it doesn't cost double what a comparable TV without Android would cost. In fact, the additional cost seems to be very close to 0.
What you have proposed has been done several times, typically with Windows Media Center. They didn't sell, for obvious reasons.
Columns in PDF's suck because they don't adapt to the window size. This in unfixable, the whole point of a PDF is that they layout is static.
HTML/CSS should do better.
But in reality the performance doesn't change, no matter which way I turn the plane.
Software (or rather, the OpenGL or Direct3D driver) has absolutely no problem rendering everything 90 degrees transformed. 2D is a different problem, but 2D is trivial in software on any remotely recent system. You can also render it on a texture and let the 3D hardware handle the problem.
Because a 200 pixel wide window sucks for pictures. Again, newspapers don't come in thin strips.
Ok I don't get this one. Are you saying that I get worse frame rates when I rotate 90 degrees so the horizon is vertical in a flight simulator, with the cockpit turned off? Or does it take longer to paint the static 2D cockpit view the other way around?
Similarly with lying on your side in an FPS, but most of them probably don't allow that.
Right, so you're after a niche product but only willing to pay mainstream prices. Good luck with that.
I'm not sure what you're on, but reading narrow columns is way faster than reading wide lines. That's why newspapers have columns. One of the many deficiencies of CSS is that it's practically impossible to a newspaper-like layout which works at any screen size (adapting the number of columns as needed).
Where did this obsession with Widescreen come from anyways? I understand for "widescreen films", but why are all monitors wide now? It's weird that it kind of slowly crept into the norm..
It's dead convenient to have one wide monitor rather than two narrow ones. Screens still aren't quite as wide as I could dream of, but we're getting there. 2:1 would be a good start.
Don't worry, the insecurity of current OS's and the ubiquity of NAT in IPv4 has forced IPv6 CPE's to block all incoming traffic. This is not quite as bad as NAT, but on the other hand it doesn't support UPNP. Good luck getting SIP to work across that.
I was in the middle of a long explanation about how your 64k is wrong, but there are much more fundamental problems with your understanding. For a starter, your understanding of RFC1918 is flawed.