Toss up between network neutrality and QoS. They could reserve a certain amount of bandwidth for initiating calls against emergency services, then once the connection is established, throw them into the normal pool of bandwidth with priority over non-emergence connections.
From my calculations of 8Kb/s, 150m people would be 1.2Tb/s, not 9.6Gb/s. Obviously I have certain assumptions on average bandiwdth used, but 8Kb isn't much. Even if you reduced it by one magnitude, you would still be at 120Gb/s.
I don't get all the hate Wayland gets. The developers of X don't even like X. If you want to take over X, go ahead, but the majority of people don't want to use X because of its performance limitations.
People who use X for features that Wayland does not support are the minority. A very vocal minority. This minority wants to impose its will over the majority.
Not only is the minority trying to tell the majority what to do, but the minority isn't even the ones who are doing the work, they're the leeches who benefit from the work of the majority.
I love how the whole GPL has breed a user base that has contempt for the developer base. If you don't like it, fork it and do it yourself. Quite your b@#ching
Know what is more expensive than fiber over distance? Copper over distance. Average operational costs of copper is about 20% higher. Even over short distances, fiber is now competitive with copper.
Population densities like those of Japan actually costs more because of the cost of installing infrastructure in already packed areas. The good news is once that fiber infrastructure is installed, the upgrade process is relatively cheap and simple.
Several objects may be sent to the GPU at the same time. Sending data to the GPU typically causes a context switch, which is expensive. DX11 gets around this by letting each worker thread write to its own command queue, then the primary context thread will notify the GPU when to read from the command queues. This effectively allows communications to the GPU to be batched instead of tossing around the GPU context between threads, which incurs a lot of overhead and is serialized.
Because the GPU is notified by the primary context, there is effectively one context switch to offload a lot of data. The driver knows where these command queues are and will read from them.
DX11 even allows the CPU to help the GPU. GPUs are great for certain types of calculations, but not as great for others. DICE had a nice blog many years ago about how they can send data to the GPU to be processed for one stage, then stream the data from the GPU as it completes, back to the CPU. The CPU then starts working on the data one one 16x16 tile at a time and streams the changes back to the GPU as each tile is completed.
They were able to keep the CPU and GPU well-loaded, while increasing efficiency and reducing memory pressure by data streaming.
They had the classic latency vs throughput issue. Because each stage was only dependent on the prior stage, they were able to keep streaming input into the engine to keep both the GPU and CPU busy. While the CPU was busy crunch special pixel shaders that ran slowly on the GPU, the GPU would be working on physics on the upcoming scene. etc etc
DX11 effectively made the rendering pipeline modular and customization, allowing data to be shifted back-and-forth, but hid the latency by allowing the stages to be done asynchronously.
OpenGL doesn't care about latency vs throughput because professional rendering does not have a latency issue, like video-games do.
MSDN Ultimate lists for $13,299 on the MSDN website. Once you pay that price, the renewal, which is every 3 years, is only $4,250, so about $1,400/year. Those are list prices. I get Pro at work and the list price is $800 and we only pay $400 every 3 years.
The Ultimate edition also includes 4 support incidents every year, where they will help you with almost any issue. You also get access to unlimited installs of every Microsoft current and past product for testing/development/presentation reasons. You also get unlimited high priority forum help. I have personally gotten quick responses on their forum from Sr developers who gave very detailed answers.
You're not paying for the IDE, you're paying for the awesome support contract.
Interstate costs about $4m/mile. I think you have a few zeros extra on your numbers. At your rates, it would cost about $19tril for our Interstate system.
What I don't understand is why states like mine have better roads and cheaper taxes than states with high amounts of road tolls. The worst roads that I have ever seen have been toll roads. And the inconvenience of having to stop at each toll is insane.
Except that you dont need your own fibre, all you need is a wavelength, so google could be just splicing cables together for say 20 households at a time, and you need only about 50m of your own fibre
Google is actually running a dedicated fiber directly from their house back to the datacenter. There is no sharing with Google Fiber, except at the uplinks.
Also, there are only a few wavelengths that actually work because IR is the best and you need certain sized guardbands.
Look up Kansas City Startup Village. It is a small business park that has Google fiber. Kansas City and Google both brag about this. There are many businesses in there that host servers and Google does not care. Google is actively encouraging Internet start ups.
Comparing apples to oranges, but the "No Servers" in the ToS is similar to the Google Docs ToS requiring that you give them copyright permissions. The legalistic nature of the USA effectively makes ISPs put in a "No Server" clause, otherwise the ISP would have nothing to fall back on if someone was using the network is an obviously abusive manner.
I have a local ISP that has a ToS that explicitly states that they are net neutral. You cannot "host servers", but they also state that they will not watch your traffic for any reason other than requested by law or by the end user and they will not traffic shape or block any traffic.
Well then, why put in a "No Server" clause in your ToS if you have no way of detecting? If you decide to host up a service and a competitor decides to DDoS your connection and the ISP goes, WTF is this 10Gb of traffic hitting our trunk?! Then they can fall-back to the ToS and say "You're running a server, stop it or get disconnected".
Without that clause, the ISP would have no way to protect itself legally in a case like that. There are probably other situations where hosting services in a certain manner can cause harm to the over-all network. Most situations probably won't cause any issues, but the ISP needs a legal safety-net.
Heck, this ISP even says "No Datacap" and "You get dedicated bandwidth". What legal recourse could this ISP have it it actively advertises that it does not block, throttle, cap, or monitor your connection while also claiming dedicated bandwidth? "No Servers".
Or another situation. Someone starts a business, and the ISP changes something on their network that affects the business's ability to host. If the ISP has a "no servers" clause, there is no question. Without the clause, the business may be able to go after the ISP for damages.
And they should be treated as such. If a warning says not to let 2-3 year-olds have something, then a teenager with a development issue probably shouldn't play with those things either.
all he wanted to do was stop his sister from being harassed by that kid.
That escalated quickly. Well, it worked, but many times I wonder if these kids are so stupid to not realize what they're doing, or fully understand that they will be ending someone's life. Either way, they should be removed from society forever because something if fundamentally wrong with them.
I've read a few case studies about modern GPON deployments in the past 5 years, and it's about $800/house passed in suburbs, and about $1500/house installed(assuming about a 50% up-take), of which about 60% of that cost is not running the fiber to the property, but the truck-roll required to do the final step of the installation. Mind you, that is based on a 32:1 ratio GPON compared to Google's dedicated fiber. I'm not sure how much of a difference that makes for capital costs.
OpenGL allows the context to be handed off to any thread, but that only means it's thread safe, but has nothing to say about its ability to work concurrently.
OpenGL does not support concurrent multi-threading as all access to the single context must be serialized.
During the beginning of the whole Kansas City Google Fiber campaign, Google said that any 3rd party ISP may make use of Google's fiber network at cost. At the end of the contract, if the city didn't want Google to still be there, Google would be willing to sell the fiber network back at cost also.
At the speeds offered, I assume not. Between Level3 Comm blogs and talking to my ISP, throttling is more expensive than adding more bandwidth. As for blocking traffic, I have not heard of anything.
They're bound to do a certain amount of monitoring, as all network admins must monitor dataflows to make sure the network is actually working, but as to the depth, like DPI, not sure. DPI at gigabit speeds on all traffic would be insanely hard, but I could possibly see focusing on certain traffic types. They could do DNS monitoring and see which domains get hit the most, but they could already do that with then OpenDNS servers, which many people already use.
Slashdot Mirror
Toss up between network neutrality and QoS. They could reserve a certain amount of bandwidth for initiating calls against emergency services, then once the connection is established, throw them into the normal pool of bandwidth with priority over non-emergence connections.
From my calculations of 8Kb/s, 150m people would be 1.2Tb/s, not 9.6Gb/s. Obviously I have certain assumptions on average bandiwdth used, but 8Kb isn't much. Even if you reduced it by one magnitude, you would still be at 120Gb/s.
If all you want are powerful features and don't give a crap about eye-candy, then use the console.
I don't get all the hate Wayland gets. The developers of X don't even like X. If you want to take over X, go ahead, but the majority of people don't want to use X because of its performance limitations.
People who use X for features that Wayland does not support are the minority. A very vocal minority. This minority wants to impose its will over the majority.
Not only is the minority trying to tell the majority what to do, but the minority isn't even the ones who are doing the work, they're the leeches who benefit from the work of the majority.
I love how the whole GPL has breed a user base that has contempt for the developer base. If you don't like it, fork it and do it yourself. Quite your b@#ching
If I can watch my progress bar, it's going too slowly.
Know what is more expensive than fiber over distance? Copper over distance. Average operational costs of copper is about 20% higher. Even over short distances, fiber is now competitive with copper.
Population densities like those of Japan actually costs more because of the cost of installing infrastructure in already packed areas. The good news is once that fiber infrastructure is installed, the upgrade process is relatively cheap and simple.
I am afraid of custom built OSI implementations. I prefer to use well tested frameworks and make sure I'm using them correctly.
QuakeSpy was great at finding Quake servers.
Several objects may be sent to the GPU at the same time. Sending data to the GPU typically causes a context switch, which is expensive. DX11 gets around this by letting each worker thread write to its own command queue, then the primary context thread will notify the GPU when to read from the command queues. This effectively allows communications to the GPU to be batched instead of tossing around the GPU context between threads, which incurs a lot of overhead and is serialized.
Because the GPU is notified by the primary context, there is effectively one context switch to offload a lot of data. The driver knows where these command queues are and will read from them.
DX11 even allows the CPU to help the GPU. GPUs are great for certain types of calculations, but not as great for others. DICE had a nice blog many years ago about how they can send data to the GPU to be processed for one stage, then stream the data from the GPU as it completes, back to the CPU. The CPU then starts working on the data one one 16x16 tile at a time and streams the changes back to the GPU as each tile is completed.
They were able to keep the CPU and GPU well-loaded, while increasing efficiency and reducing memory pressure by data streaming.
They had the classic latency vs throughput issue. Because each stage was only dependent on the prior stage, they were able to keep streaming input into the engine to keep both the GPU and CPU busy. While the CPU was busy crunch special pixel shaders that ran slowly on the GPU, the GPU would be working on physics on the upcoming scene. etc etc
DX11 effectively made the rendering pipeline modular and customization, allowing data to be shifted back-and-forth, but hid the latency by allowing the stages to be done asynchronously.
OpenGL doesn't care about latency vs throughput because professional rendering does not have a latency issue, like video-games do.
MSDN Ultimate lists for $13,299 on the MSDN website. Once you pay that price, the renewal, which is every 3 years, is only $4,250, so about $1,400/year. Those are list prices. I get Pro at work and the list price is $800 and we only pay $400 every 3 years.
The Ultimate edition also includes 4 support incidents every year, where they will help you with almost any issue. You also get access to unlimited installs of every Microsoft current and past product for testing/development/presentation reasons. You also get unlimited high priority forum help. I have personally gotten quick responses on their forum from Sr developers who gave very detailed answers.
You're not paying for the IDE, you're paying for the awesome support contract.
Their free tier is asymmetric, their only other paid tier is 1Gb symmetric. Talk about splitting hairs.
Interstate costs about $4m/mile. I think you have a few zeros extra on your numbers. At your rates, it would cost about $19tril for our Interstate system.
What I don't understand is why states like mine have better roads and cheaper taxes than states with high amounts of road tolls. The worst roads that I have ever seen have been toll roads. And the inconvenience of having to stop at each toll is insane.
Too dense is also bad. I forget the optimal density, but it is close to the average USA city.
Except that you dont need your own fibre, all you need is a wavelength, so google could be just splicing cables together for say 20 households at a time, and you need only about 50m of your own fibre
Google is actually running a dedicated fiber directly from their house back to the datacenter. There is no sharing with Google Fiber, except at the uplinks.
Also, there are only a few wavelengths that actually work because IR is the best and you need certain sized guardbands.
Look up Kansas City Startup Village. It is a small business park that has Google fiber. Kansas City and Google both brag about this. There are many businesses in there that host servers and Google does not care. Google is actively encouraging Internet start ups.
Comparing apples to oranges, but the "No Servers" in the ToS is similar to the Google Docs ToS requiring that you give them copyright permissions. The legalistic nature of the USA effectively makes ISPs put in a "No Server" clause, otherwise the ISP would have nothing to fall back on if someone was using the network is an obviously abusive manner.
I have a local ISP that has a ToS that explicitly states that they are net neutral. You cannot "host servers", but they also state that they will not watch your traffic for any reason other than requested by law or by the end user and they will not traffic shape or block any traffic. Well then, why put in a "No Server" clause in your ToS if you have no way of detecting? If you decide to host up a service and a competitor decides to DDoS your connection and the ISP goes, WTF is this 10Gb of traffic hitting our trunk?! Then they can fall-back to the ToS and say "You're running a server, stop it or get disconnected".
Without that clause, the ISP would have no way to protect itself legally in a case like that. There are probably other situations where hosting services in a certain manner can cause harm to the over-all network. Most situations probably won't cause any issues, but the ISP needs a legal safety-net.
Heck, this ISP even says "No Datacap" and "You get dedicated bandwidth". What legal recourse could this ISP have it it actively advertises that it does not block, throttle, cap, or monitor your connection while also claiming dedicated bandwidth? "No Servers".
Or another situation. Someone starts a business, and the ISP changes something on their network that affects the business's ability to host. If the ISP has a "no servers" clause, there is no question. Without the clause, the business may be able to go after the ISP for damages.
And they should be treated as such. If a warning says not to let 2-3 year-olds have something, then a teenager with a development issue probably shouldn't play with those things either.
CPSC has received 54 reports of children and teens ingesting this product, with 53 of these requiring medical interventions.
Sounds like Darwinism in action. Young children, I can understand, but teens?
all he wanted to do was stop his sister from being harassed by that kid.
That escalated quickly. Well, it worked, but many times I wonder if these kids are so stupid to not realize what they're doing, or fully understand that they will be ending someone's life. Either way, they should be removed from society forever because something if fundamentally wrong with them.
Wire 15% of the USA or have Google waste money on patents.. oh, sorry, wrong topic. /cry
I've read a few case studies about modern GPON deployments in the past 5 years, and it's about $800/house passed in suburbs, and about $1500/house installed(assuming about a 50% up-take), of which about 60% of that cost is not running the fiber to the property, but the truck-roll required to do the final step of the installation. Mind you, that is based on a 32:1 ratio GPON compared to Google's dedicated fiber. I'm not sure how much of a difference that makes for capital costs.
OpenGL allows the context to be handed off to any thread, but that only means it's thread safe, but has nothing to say about its ability to work concurrently.
OpenGL does not support concurrent multi-threading as all access to the single context must be serialized.
During the beginning of the whole Kansas City Google Fiber campaign, Google said that any 3rd party ISP may make use of Google's fiber network at cost. At the end of the contract, if the city didn't want Google to still be there, Google would be willing to sell the fiber network back at cost also.
At the speeds offered, I assume not. Between Level3 Comm blogs and talking to my ISP, throttling is more expensive than adding more bandwidth. As for blocking traffic, I have not heard of anything.
They're bound to do a certain amount of monitoring, as all network admins must monitor dataflows to make sure the network is actually working, but as to the depth, like DPI, not sure. DPI at gigabit speeds on all traffic would be insanely hard, but I could possibly see focusing on certain traffic types. They could do DNS monitoring and see which domains get hit the most, but they could already do that with then OpenDNS servers, which many people already use.
but then maybe a family of 4 that is doing that *should* be paying more than I am.
And people who use the library more should pay more in taxes.