Yes, but some matter can get very cold and become very dark, making it absorb visible light than it emits, making it a good source of blocking light, but not emitting. For example. There is a huge cloud of gas that is colder than the cosmic background. While it is absorbing light, it is also expanding, which cools it. The cooling from expansion is greater than heating from incoming radiation.
We've been bouncing microwaves around cavities for a long time, and we've never come up with something like this.
I would stay away from arguments like this. Turns out many of the answers to "difficult questions" are right under our noses, sometimes literally. For example. We've been working on fighting MRSA for a long time. So happens that bacteria in our noses and gut secrete anti-biotics that have thus far killed MRSA with a 100% success rate in rats.
Doing something for a long time is a pretty bad metric of useful experience or mastery of a subject. You need someone who thinks differently.
Dark matter is the name of the "going on". If you recognize that something is going on, then you imply that "Dark Matter" is real effect. To say Dark Matter is not real is to say the "going on" is not real.
Effort is a useless metric unless compared against someone mastered in the art. If a master would have found the issue effortlessly, assuming no hindsight, then the company did not make a good-faith "effort". They probably hired someone with "security skills" and checked off a box somewhere saying they did a security review, assuming that even happened.
Humans are just computers. I'm confused on why you are confused about the discussion. Treating humans specially is just a current silliness. The example of past silliness is because the past was irrational. Rationality is universal and timeless. If something is "rational" today but not tomorrow, that's because it was never rational to begin with.
Slavery is a form of high interest debt. Eventually you pay for it. Like a 15% APR credit-card that you only have to pay down once a century or two. You may make the first few payments, but one of them is going to do you in.
Close but not quite. Residential customers get a discount because residential users as a group have a different bandwidth profile. It's not even just this simple. ISPs costs are typically based on peak usage, which is driven by residential customers. Bandwidth used outside of normal residential peak hours is virtually free. Around here, businesses actually pay the same residential users for the same service.
Key words there, "same service". My ISP makes no real distinction between business and residential. Both can get static IP blocks for cheap, uncapped dedicated bandwidth, individual strands of fiber back to the CO. If you want an SLA, prepare to pay through the nose.
That is not the definition of "shaping", that is Cisco's definition for their own internal terminology. Regardless of what you want to call it, I can control the amount of bandwidth a flow or group of flows can use regardless of direction (ingress/egress), assuming they respond to normal loss, marked, or delayed packet. Most people calling this "shaping bandwidth", but you can call it whatever you want.
Remember that you're trading that extra latency for lower probability of dropped packets.
Not once you've gotten into the "bloated" range of buffer sizes. Increased latency from large buffers also increases the latency to signal to the sender that the route is congested. The sender will spend more time sending packets that will ultimately just get dropped. If the latency was lower, the sender would have known sooner to reduce its rate. Latency and loss go hand-in-hand once you get into unnaturally large buffers. I'm not sure the exact recommend buffer size, but I think it's around 10ms of the bandwidth. Many people are seeing 1,000ms+, which is 2 orders magnitudes above optimal.
+1 this up. Nailed it, and not the meme type. Most people, including network admins with years of working with QoS, are incapable of setting up QoS correctly, and only think they've set it up correctly not because of theoretical correctness but because they cannot even think of the edge cases to get empirical tests..
At some point in your life, you should realize that "no one" mean very very rarely. Absolutes are never absolute. I question if my last statement makes any sense.
This is actually normal for any congestion control algorithm that uses only packetloss to signal congestion. TCP? It keeps sending data until the buffer fill and drops "a packet", but we really know FIFO taildrop buffers drop bunches of packets. Then TCP backs off. But wait, there's more! You have many TCP flows going over the connection, so they are all fluctuation, keeping the buffer either in a state of steady full, which causes high latency and lots of dropped packets, or wildly swinging between empty an full because of global synchronization.
The only real way to solve this is to timestamp packets as they enter a buffer and drop the ones that are too old.
You don't have to timestamp them to get the same effect. Codel and RED both effectively use time without timestamping. But yes, the "tracking time" is pretty much the only way.
The article talks about shaping download, which isn't possible at the endpoint. The traffic is already there and you have to deal with it. Dropping it will create retransmits for TCP and make the problem worse.
Wrong. Dropped packets signal congestion. If you don't signal congestion, the congestion will only get worse. You eventually have to drop a packet. The sooner you drop the packet after congestion has started, the less the congestion will be. The flip side is if you signal too early, you lose effective bandwidth. I shape my download and it has caused my average to go up because it stabilizes the flows.
With normal fifo buffers, once the buffer is full, you get a burst of lost packets. This is much worst than dropping a single packet earlier.
Buffering can reduce latency, especially under heavy load, by better bandwidth utilization
You have no idea what you're talking about. Buffering is one of the main causes of latency. Ever see a 1,000ms ping? That's not because the speed of light is too slow, that's because there is a backlog of packets in the buffer. With the speed of light through fiber, no one should ever see a ping above 300ms to anywhere in the world. The highest ping I see from Midwest USA to Australia, India, or China is about 220ms.
Buffers are not inherently bad, but "bufferbloat" is because buffers are too large. Too large of buffers actually reduce throughput because TCP takes longer to respond to changes in congestion. Even worse is when bufferbloat starts to get up into the 3second range, yes seconds not milliseconds, TCP treats it as a lost packet and resends the data. I regularly see bloated Linux ISO seeders with 2k-4kms pings resending nearly 50% of their packets, most of which were not actually lost but only highly delayed.
Good anti-bufferbloat AQMs like fq_Codel and Cake increase effective bandwidth, while isolating light traffic from heavy traffic and keeping latency almost idle-link low. Want 10ms pings while paying games and downloading/uploading torrents, I have that already.
One of the FreeBSD foundation servers started off as something like FreeBSD 2.0 32bit, and has been getting in-place upgrades since. It was last announced was FreeBSD 9.3 64bit and running executables that are nearly 20 years old and they no longer have the source-code for. I thought the idea of in-place upgrades has been solved for decades.
What does Google actually do that running your own web crawler doesn't? Ah, turn self-hosting into a centralized ad platform controlled by one company.
Github doesn't plate, it complements. It makes it easier to discover and collaborate. It does not change how git works, just everything around git.
Most people have delicate egos and only like to have others who know how to stroke it around them. Aka, being "social". Of course you don't want to be a party-pooper, but I prefer people who just tell me what's on their mind rather than act fake to blend in.
The first being, courts could simply say "these contracts are unenforceable"
Not they can't. It's up to courts to enforce the law. If the law doesn't say one way or the other and someone signed a contract, then it's legally binding. I assume you agree someone can sign away their freedom. Then everyone could add a clause that you will become their slave. Sure, you can't force anyone to sign it, but businesses can collude to create it as something that is standard. Nothing you can do.
Just you wait until stacked memory goes mainstream. It won't be long before a core has 64GiB of multi-TiB/s memory integrated directly above. High end GPUs will be the first to get, but I see CPUs going the same direction not long after. I would gladly give up expandable memory to have huge amounts of integrated fast memory.
Slashdot Mirror
Yes, but some matter can get very cold and become very dark, making it absorb visible light than it emits, making it a good source of blocking light, but not emitting. For example. There is a huge cloud of gas that is colder than the cosmic background. While it is absorbing light, it is also expanding, which cools it. The cooling from expansion is greater than heating from incoming radiation.
Momentum is mass times velocity
That is an over-simplification. Light has momentum yet has no mass.
We've been bouncing microwaves around cavities for a long time, and we've never come up with something like this.
I would stay away from arguments like this. Turns out many of the answers to "difficult questions" are right under our noses, sometimes literally. For example. We've been working on fighting MRSA for a long time. So happens that bacteria in our noses and gut secrete anti-biotics that have thus far killed MRSA with a 100% success rate in rats.
Doing something for a long time is a pretty bad metric of useful experience or mastery of a subject. You need someone who thinks differently.
Dark matter is the name of the "going on". If you recognize that something is going on, then you imply that "Dark Matter" is real effect. To say Dark Matter is not real is to say the "going on" is not real.
Great point about the age. It does bring up the subject of how long security support should be offered in the way of updates and patches.
Effort is a useless metric unless compared against someone mastered in the art. If a master would have found the issue effortlessly, assuming no hindsight, then the company did not make a good-faith "effort". They probably hired someone with "security skills" and checked off a box somewhere saying they did a security review, assuming that even happened.
Reproduction and intelligence are inversely related. This is not good for us.
Humans are just computers. I'm confused on why you are confused about the discussion. Treating humans specially is just a current silliness. The example of past silliness is because the past was irrational. Rationality is universal and timeless. If something is "rational" today but not tomorrow, that's because it was never rational to begin with.
Slavery is a form of high interest debt. Eventually you pay for it. Like a 15% APR credit-card that you only have to pay down once a century or two. You may make the first few payments, but one of them is going to do you in.
Close but not quite. Residential customers get a discount because residential users as a group have a different bandwidth profile. It's not even just this simple. ISPs costs are typically based on peak usage, which is driven by residential customers. Bandwidth used outside of normal residential peak hours is virtually free. Around here, businesses actually pay the same residential users for the same service.
Key words there, "same service". My ISP makes no real distinction between business and residential. Both can get static IP blocks for cheap, uncapped dedicated bandwidth, individual strands of fiber back to the CO. If you want an SLA, prepare to pay through the nose.
That is not the definition of "shaping", that is Cisco's definition for their own internal terminology. Regardless of what you want to call it, I can control the amount of bandwidth a flow or group of flows can use regardless of direction (ingress/egress), assuming they respond to normal loss, marked, or delayed packet. Most people calling this "shaping bandwidth", but you can call it whatever you want.
The Mythical Man Month. Most people follow the antithesis of this book.
Remember that you're trading that extra latency for lower probability of dropped packets.
Not once you've gotten into the "bloated" range of buffer sizes. Increased latency from large buffers also increases the latency to signal to the sender that the route is congested. The sender will spend more time sending packets that will ultimately just get dropped. If the latency was lower, the sender would have known sooner to reduce its rate. Latency and loss go hand-in-hand once you get into unnaturally large buffers. I'm not sure the exact recommend buffer size, but I think it's around 10ms of the bandwidth. Many people are seeing 1,000ms+, which is 2 orders magnitudes above optimal.
+1 this up. Nailed it, and not the meme type. Most people, including network admins with years of working with QoS, are incapable of setting up QoS correctly, and only think they've set it up correctly not because of theoretical correctness but because they cannot even think of the edge cases to get empirical tests..
In a lot of places
At some point in your life, you should realize that "no one" mean very very rarely. Absolutes are never absolute. I question if my last statement makes any sense.
If it's constantly filling up and evicting
This is actually normal for any congestion control algorithm that uses only packetloss to signal congestion. TCP? It keeps sending data until the buffer fill and drops "a packet", but we really know FIFO taildrop buffers drop bunches of packets. Then TCP backs off. But wait, there's more! You have many TCP flows going over the connection, so they are all fluctuation, keeping the buffer either in a state of steady full, which causes high latency and lots of dropped packets, or wildly swinging between empty an full because of global synchronization.
The only real way to solve this is to timestamp packets as they enter a buffer and drop the ones that are too old.
You don't have to timestamp them to get the same effect. Codel and RED both effectively use time without timestamping. But yes, the "tracking time" is pretty much the only way.
The article talks about shaping download, which isn't possible at the endpoint. The traffic is already there and you have to deal with it. Dropping it will create retransmits for TCP and make the problem worse.
Wrong. Dropped packets signal congestion. If you don't signal congestion, the congestion will only get worse. You eventually have to drop a packet. The sooner you drop the packet after congestion has started, the less the congestion will be. The flip side is if you signal too early, you lose effective bandwidth. I shape my download and it has caused my average to go up because it stabilizes the flows.
With normal fifo buffers, once the buffer is full, you get a burst of lost packets. This is much worst than dropping a single packet earlier.
My response to this, it's not even wrong.
Buffering can reduce latency, especially under heavy load, by better bandwidth utilization
You have no idea what you're talking about. Buffering is one of the main causes of latency. Ever see a 1,000ms ping? That's not because the speed of light is too slow, that's because there is a backlog of packets in the buffer. With the speed of light through fiber, no one should ever see a ping above 300ms to anywhere in the world. The highest ping I see from Midwest USA to Australia, India, or China is about 220ms.
Buffers are not inherently bad, but "bufferbloat" is because buffers are too large. Too large of buffers actually reduce throughput because TCP takes longer to respond to changes in congestion. Even worse is when bufferbloat starts to get up into the 3second range, yes seconds not milliseconds, TCP treats it as a lost packet and resends the data. I regularly see bloated Linux ISO seeders with 2k-4kms pings resending nearly 50% of their packets, most of which were not actually lost but only highly delayed.
Good anti-bufferbloat AQMs like fq_Codel and Cake increase effective bandwidth, while isolating light traffic from heavy traffic and keeping latency almost idle-link low. Want 10ms pings while paying games and downloading/uploading torrents, I have that already.
One of the FreeBSD foundation servers started off as something like FreeBSD 2.0 32bit, and has been getting in-place upgrades since. It was last announced was FreeBSD 9.3 64bit and running executables that are nearly 20 years old and they no longer have the source-code for. I thought the idea of in-place upgrades has been solved for decades.
What does Google actually do that running your own web crawler doesn't? Ah, turn self-hosting into a centralized ad platform controlled by one company.
Github doesn't plate, it complements. It makes it easier to discover and collaborate. It does not change how git works, just everything around git.
Most people have delicate egos and only like to have others who know how to stroke it around them. Aka, being "social". Of course you don't want to be a party-pooper, but I prefer people who just tell me what's on their mind rather than act fake to blend in.
The first being, courts could simply say "these contracts are unenforceable"
Not they can't. It's up to courts to enforce the law. If the law doesn't say one way or the other and someone signed a contract, then it's legally binding. I assume you agree someone can sign away their freedom. Then everyone could add a clause that you will become their slave. Sure, you can't force anyone to sign it, but businesses can collude to create it as something that is standard. Nothing you can do.
Just you wait until stacked memory goes mainstream. It won't be long before a core has 64GiB of multi-TiB/s memory integrated directly above. High end GPUs will be the first to get, but I see CPUs going the same direction not long after. I would gladly give up expandable memory to have huge amounts of integrated fast memory.