The smart kids that I knew rarely studied and got to skip most of the "normal" classes and instead do special projects, then skip their capstone because their special projects were above and beyond the requirements of the capstone. Some of them got to help teach classes and get credit for the class and not have to do a lick of homework for the class.
What about those of us who understand and can use math, but cannot do math? I can use math in that I know what operations need be done to solve a problem, but I myself cannot do the operations because of a learning disability. Another example is I was talking to my one of my ex-biology teachers about some new discovery in metabolism. We went deep into theory. At the end of it she asked me if I was a grad student because she didn't remember me in her classes, but I looked familiar. I told her I failed Bio 101. She as flabbergasted that I failed with all of the understanding of difficult theories that I just showed her. I told her, I have a learning disability that I have extreme difficulty with rote memorization, but I have a deep understanding of abstract concepts.
As more research is being done, more and more is showing an inverse relationship between knowledge and understanding. Of course you need basic knowledge, but beyond basic knowledge, the more you remember, the less you understand. The brain seemingly doesn't like to do both. Either you have strong reasoning skills or you have strong rote memory skills. Rote memory can include stuff like experience. which can make someone seemingly have good reasoning skills, but only because they face-planted in the past and remember the painful experience, when the failure could have been avoided with better reasoning.
There was another/. from a long while back that said walking every 1.5-2hr for 10-15min showed reduced heart issues over those who exercised more and did proper cardio, but only after work. If there's anything I learned, it's moderation. You don't need a lot of exercise, you just need to get some here and there throughout the day.
Talent doesn't follow money, it follows opportunity, which is correlated with money. If you think increasing the wages increases talent, you're waaaayyy off. I've seen companies paying talent nearly $1mil/year to work remotely, and lost them when they took the fun out of work. Later that person turns up working for $100k in some start-up.
At first I felt the same about what he said about "how it's done" not mattering, but I thought about it and later assumed he meant "an action, in and of itself, does not matter, it's choosing the right action". Kind of a cargo-cult take on what they may have meant.
But..."knows the how to produce outputs rather than worry about tinkering with how things work" is a very dangerous thing indeed. If you don't know how something works, how can you even possibly know you did the correct thing? It's logically impossible. I deal day-in-and-out with people who made solution to mask the problem, not fix the problem, because they didn't take the time to "tinker' and figure out how something works. Rewarding people with getting immediate results over doing something correctly is how technical debt skyrockets. It just externalizes the costs into the future and for other people.
I deal with managers who tell me technical debt can be good if managed correctly because it allows you to do something sooner. My argument is that is correct, if you properly measure the debt, but most of the time they're not taking out mortgages and flipping houses, they're taking out pay-day loans on money-pits. That week they saved on a month long project set us back 2 months and tens of man-years of work in less than a half a year's time.
My personal take on problem solving is if you don't know how something should work and how it does work, then you have no idea what you're doing and you're just throwing crap at the wall and hoping it sticks. Same thing goes for how something doesn't work. How something works is the complement set of how something doesn't work. By definition, if you know one, you must know the other. Most of the time that something fails and I ask why it's failing, the programmer has no idea why it's failing. By my definition, if they don't know why it's failing, they could not have known how it works in the first place. If they don't know how it works, then they have no idea what they wrote in the first place. Just throwing crap at the wall.
When my code fails, I nearly always have an immediate theory as to why it's failing and most of the time my theory is perfectly correct or nearly so. Heck. Even when helping other people with their projects and figuring out why their code is not working correctly, my intuited guesses are better than their educated and informed theories. When I make these kinds of mistakes myself, I will fester on my mistake for weeks. I am my own worst critic and I hold myself to a high standard.
It's more rewarding to produce a system that is difficult to break. Breaking a sandcastle is fun, but building a sand castle that can't be broken is even more fun. When it is eventually broken, you learn from your mistake in lack of creativity.
I hate it when programmers blame the libraries they use. Never write a line of code or use a library unless you know what it's doing and how it's doing it. When I use a library, I'm hitting up github and reading their code.
Net neutrality is about the delivery mechanism being fair for all sources for a class of data. And "fair" being defined as something like "not interfering" or "not preferring". Peering with someone is not "preferring" because it does not manipulate the transmission in any way, it only brings the source "closer".
All of the big companies have found out how difficult it is to hire top talent by offering truck loads of money. If I'm in the top 0.1% of where I live and I love where I live, why would I care to get more money? There's actually been studies about offering more money and results. Offering too little is detrimental, but offering too much can be nearly as bad. Most people just want to be paid enough to do what they want to do an feel secure in their financial situation.
In my limited experience, skill and ability are inversely related. Highly skilled people are people who spent a lot of time doing a few things. They don't understand anything outside of the few things they do. In the end, they have this great skill of hammering nails, but they can't tell the difference between a screw an a nail.
Quite a bit of research is starting to show that in tech, masters are rarely very skillful and experienced workers are rarely masterful. Mastery comes from a deep understanding, which can be acquired independently of knowledge or experience. In many other professions, mastery and experience are highly correlated, but not so with highly cognitive and creative professions. A surgeon needs to physically experience the difficulty of suturing with blood soaked hands, but a programmer does not need to experience not validating input from an un-trusted source to know something bad will eventually happen. In the programmer's case, most of their issues can be 100% reasoned about, no experience required.
In problem domains where reasoning is almost the only thing you need to do to "learn", people who learn by thinking instead of doing can learn hundreds or thousands of times faster. The saying is it takes 10,000 hours to become a master. That is a generalization. It's 1,000-100,000 hours depending on the person and for domains where practice matters. Someone who becomes a master in 1,000 hours does so via deliberate practice, while the person who takes 100,000 hours via accidental practice. For domains where practice has virtually zero correlation, like programming, it becomes even more pronounced with more like 100-1,000,000 depending on the person. One person becomes a master in 2 weeks and another takes 350 years.
Not to mention that tech skills have a half-life of around 3-5 years. If it takes you more than 1 year to master something, it'll be useless by the time you master it. Domain transference is an ability of Abstract Reasoning. If you have strong Abstract Reasoning, you don't learn skills, you learn patterns. When you need to learn a new skill, you don't need to learn an entirely new skill, you abstract it into its fundamental patterns and transfer skill/knowledge from another domain.
I haven't had to learn a new tech skill in decades. Most of my tech skills I learned before I even owned a computer. The only skills I have remaining are non-tech related, like cleaner code, working as a team, communications, documentation, and other soft skills. Tech skills have been mostly unchanging. Just learning new tools for the same old problems that were solved back in the '60s.
I should add my family is a fun-loving group of witty people who love a good practical joke and a nice relaxing beer or two while hanging out with friends. When it comes to work, keep up or get out of the way. We do go well out of our way to help people and we enjoy being part of a team when the opportunity arises. We hate jerks and we loath drama.
We do have our black sheep. I was told my one uncle used to be normal, but then got into drugs for a decade. No one really knew where he was most of the time because he was coked out of his skull. Even after he got done with his drug spree because he was messed up in the head, he tested so well, he got offered several free rides through a few Ivy League Universities. My grandma had to turn them down because she had to take care of him. He may have lost most of his grip on reality, but he was still freaking smart, which was a scary mix when he started talking about the devil taking over the world. He kept family reunions lively.
In my blood family, my generation is the first to go to college and nearly everyone one of my cousins were valedictorian or dean's list. Several of my cousins are now millionaires, and we all come from poor families. Most of my aunts and uncles are making about $20k/year, but their children are now $100k-$1mil in very low income areas. We like to stay close to home. And I have about 50 cousins. Pretty much everyone of my cousins are business owners, specialized medical doctors, or some form of Computer Science. Many of whom get personally invited for talks in their specialties around the nation. They don't submit papers to see if they get accepted, they get personally asked to host a talk.
My mom only has a GDE, and is a 55 year old secretary. She only recently got a computer and uses it maybe a few times a month. I can talk to her about optimizing distributed computing better than the CS graduates that I have to work with. Even my brother was a sophomore in college when he was leading 6 different graduate research teams at the same time. They had to take him off the teams and place him on his own because he was doing all of the work. The graduate students could not keep up. During his junior year in college, he had research Universities that dealt in a range of specializations from Super Computing to AI research, vying for him for a paid internship. In the end he took the super computing internship. Got paid $10k for the summer with room and board for free. Plus exclusive use of the entire super computer, normally reserved for PHDs. during his senior year, he did a one man $3mil datacenter project, also got to work with the FBI in enhancing their security, and taught several classes at the Uni.
Pretty much everyone in my family can talk to each other as peers, even if we're in different specialties. We have difficulty working as peers with most people, but we rarely need to seeing we have few peers.
No one has called me an "arrogant asshole" because I enjoy being around happy people and see no reason to piss people off. My entire life while in school I was considered stupid until they found out I was gifted and just think differently. I did not like being judged, so I don't like judging others.
Don't go thinking everybody's life experience is the same as yours. If *you* hit the jackpot and got a great job straight out of college with no experience, don't assume experience is therefore worthless for anyone and everyone.
That wasn't what I was going for. By definition, Abstract Reasoning is the ability to solve problems without experience or knowledge. If the discussion is about domains like Computer Science or general programming where strong Abstract Reason is considered an absolute must, then experience and knowledge are by definition moot. Part of the definition of Abstract Reasoning is the ability to synthesize knowledge. If you need knowledge, make it, don't learn it. Of course this only applies to the technical aspects.
If *you* hit the jackpot
Then nearly everyone in my blood family hit the jackpot, on their own, with no help from their parents or the rest of the family.
Experience is nearly worthless. My brother was making $100k/year working 20 hours per week while in college, placing him in the top 0.1% of the local household income. Then he graduated, immediately with a job for $120k/year plus great benefits. In less than a month, he was promoted to the head of his department. He had the job lined up for over a year. They didn't care how long it took my brother to graduate, they were saving him a position.
They are not paying top dollar for my brother's experience, but his ability to solve difficult problems. He was lucky that he got a local job. My professors told me they knew people in high places at some of the biggest companies and were willing to get me a good job, but I turned them down because I would have to leave the state and be away from family.
Experience teaches you how to not make the same mistakes, but it does not teach you how to not make mistakes in the first place. Experience and knowledge are the worst ways of learning problem solving.
The poor people I've seen have iPhones, just a model or two older. I can get an iPhone 5 with an unlimited dataplan for $20/m cheaper than a basic cell phone.
After looking more into it, it seems that rule of thumb is great for most situations, and probably especially most videos. For relatively static images, the effective acuity of the eye goes up as it re-scans the image. Even at 294 pixels per inch at arm's length, each pixel is going to stimulate around a hundred nerves. That means the underlying raw optical sensory of the eye is still 100x higher than the rule of thumb, we just don't get to directly perceive it for many reasons related to pre/post perceptual filter and pruning of visual data.
I'm getting the feeling of apples to oranges, similar to the "FPS" debate about how fast the eyes can see. There are several dimensions in which the eye can be measured, and normalizing it down to a single dimension of "frames per second" does not properly capture the complex situation.
I downloaded a PDF recently onto my S7 where the font size was quite small. So small that the capital "T" is about the size of the dot in "i" on my 23" 1080p monitor at font size 8. I was able to read this document without zoom at slightly over 12" away. Capital letters were only slightly larger than a few pixels on my monitor, yet I could read them with only minor strain and slightly reduced speed.
I was reading an interview with a VR specialist that said VR won't be able to do photorealistic rendering until somewhere between 8Mp and 16Mp per eye. While 720p is technically good enough from the rule of thumb, the rule of thumb breaks down on certain extreme edge cases. If we want VR to look like the real world, we're going to need between 16 and 32 Megapixels, at 300-1000 frames per second and about 16 bits of color depth for contrast.
Sounds like big numbers until you put them into perspective. Infrastructure and transit bandwidth is about 1%-3% of an established ISP's overall operating expenses. Network equipment and running the fiber constitutes only about 20% of the up-front costs. About 95% of the operating costs of an ISP are Marketing and customer support. One support truck-roll will cost an ISP nearly all of their profits for the customer for a year. Spending 15 minutes on the phone with a customer pays for nearly an entire year's worth of transit bandwidth and network upgrades.
The biggest hurdle to becoming and ISP is not the money, it's the red tape. Even if you get the venture capital, you will be held up by the local monopoly by denying you access or keeping you in court. ISPs are not federally recognized to have access to right-of-ways, only Cable and Telcoms, of which ISPs are neither. The easiest way to gain access to right-of-ways is to become a Cable company or Telcom, but then you immediately get heavily regulated and much of your profits go out the window.
It's nearly impossible to find ISPs that are only ISPs for these reasons. Our fastest internet is in the sticks where no big ISPs want to go. I can get a dedicated 500/500 fiber internet at a farm in the middle of no where for $100/m from a local private ISP with zero government subsidies. Or I can go to the local city and get 60/4 cable from Charter for $100/m, and rarely get 60 during peak hours and fight through the peak time lag and loss.
And boy do I mean dedicated. You will get your full speed 24/7 to nearly the entire world while having less than 1ms of jitter and less than 0.00001% average loss. I can go weeks without losing a single ICMP packet to Germany from Midwest USA or more than a brief 4ms flutter. My min ping is equal to my avg ping and my std-dev is less than 0.1ms from the average. Even under DDOS testing, my ping is never more than 20ms greater than the norm, but my loss goes sky high.
Even 384kbps mp3 sounds like crap. May not be able to tell the difference between 192 and 128, but I can easily tell the difference between lossless and mp3. Mp3 compression can cause my physical pain in my ears. I have no idea how this does not bother other people. I will get what feels and sounds and feels like swimmer's ear after a few minutes then my ears will start ringing and give me physical pain like an ear infection after a few more minutes. Ogg does not do this and of course not flac.
Most DJs I know use flac, but every so often an mp3 makes its way in and I either have to leave the room for a bit or ask the DJ to change the music. I typically just leave. If it's a DJ I'm paying for, I just tell them ahead of time not to play any mp3s.
Eyes have have a definition of about 6 megapixels in the center 5%. Perception filters out most of the details because they're not needed, but any unexpected visual data will stand out, like compression artifacts. If all of your vision was as good as the center, we'd need about 300Mp-400Mp per eye. As it stands, 75% of our visual acuity is in 5% of our vision. A non-uniform distribution with an extremely high standard deviation makes taking about averages meaningless.
DevOps is a cultural mindset, not a position.The job of an admin is to run a server. If you go to the cloud, with proper devops, your admins are pretty much out of a job.
QoS is only effective when there is not enough bandwidth. When pushed against the limitations of technology, QoS is expensive. Would you rather have 120Gb/s with QoS or 600Gb/s with no QoS. These are the choices that must be made.
A la carte works best when you don't want every channel under the Sun. If you want it all, sticking with your cable provider is probably your best bet. The point is most people don't want everything. At least the bills are more predictable. For a while on Charter, my bill was going up $2-$3 every month or two with no explanation while I was in contract. After a year of it, I had my wife call them up and start a fire storm. My wife can be down right evil on the phone, but only because she knows you have the play the game to get anywhere.
I just don't want any billing shenanigans. If I can get that while paying within 10%-20% for exactly the channels I want, that's a win.
Quantum computers change symmetric key encryption from O(2^n) to O(2^(n/2)), saying nothing about the time take for each operation. But quantum computer can reverse certain types of private key crypto in the same number of operations as it takes to encrypt it, rendering those cryptos completely useless. Ignoring that currently adding more bits to a quantum computer is incredibly hard. Certain aspects of these algorithms could be scaled up to to where the quantum computer could not hold all of the state. RSA scales very poorly in size and people hate increasing the key size.
There is working being done to create an asymmetric crypto that has no quantum weaknesses. I think Lattice encryption is one such.
Slashdot Mirror
The smart kids that I knew rarely studied and got to skip most of the "normal" classes and instead do special projects, then skip their capstone because their special projects were above and beyond the requirements of the capstone. Some of them got to help teach classes and get credit for the class and not have to do a lick of homework for the class.
What about those of us who understand and can use math, but cannot do math? I can use math in that I know what operations need be done to solve a problem, but I myself cannot do the operations because of a learning disability. Another example is I was talking to my one of my ex-biology teachers about some new discovery in metabolism. We went deep into theory. At the end of it she asked me if I was a grad student because she didn't remember me in her classes, but I looked familiar. I told her I failed Bio 101. She as flabbergasted that I failed with all of the understanding of difficult theories that I just showed her. I told her, I have a learning disability that I have extreme difficulty with rote memorization, but I have a deep understanding of abstract concepts.
As more research is being done, more and more is showing an inverse relationship between knowledge and understanding. Of course you need basic knowledge, but beyond basic knowledge, the more you remember, the less you understand. The brain seemingly doesn't like to do both. Either you have strong reasoning skills or you have strong rote memory skills. Rote memory can include stuff like experience. which can make someone seemingly have good reasoning skills, but only because they face-planted in the past and remember the painful experience, when the failure could have been avoided with better reasoning.
There was another /. from a long while back that said walking every 1.5-2hr for 10-15min showed reduced heart issues over those who exercised more and did proper cardio, but only after work. If there's anything I learned, it's moderation. You don't need a lot of exercise, you just need to get some here and there throughout the day.
Talent doesn't follow money, it follows opportunity, which is correlated with money. If you think increasing the wages increases talent, you're waaaayyy off. I've seen companies paying talent nearly $1mil/year to work remotely, and lost them when they took the fun out of work. Later that person turns up working for $100k in some start-up.
At first I felt the same about what he said about "how it's done" not mattering, but I thought about it and later assumed he meant "an action, in and of itself, does not matter, it's choosing the right action". Kind of a cargo-cult take on what they may have meant.
But..."knows the how to produce outputs rather than worry about tinkering with how things work" is a very dangerous thing indeed. If you don't know how something works, how can you even possibly know you did the correct thing? It's logically impossible. I deal day-in-and-out with people who made solution to mask the problem, not fix the problem, because they didn't take the time to "tinker' and figure out how something works. Rewarding people with getting immediate results over doing something correctly is how technical debt skyrockets. It just externalizes the costs into the future and for other people.
I deal with managers who tell me technical debt can be good if managed correctly because it allows you to do something sooner. My argument is that is correct, if you properly measure the debt, but most of the time they're not taking out mortgages and flipping houses, they're taking out pay-day loans on money-pits. That week they saved on a month long project set us back 2 months and tens of man-years of work in less than a half a year's time.
My personal take on problem solving is if you don't know how something should work and how it does work, then you have no idea what you're doing and you're just throwing crap at the wall and hoping it sticks. Same thing goes for how something doesn't work. How something works is the complement set of how something doesn't work. By definition, if you know one, you must know the other. Most of the time that something fails and I ask why it's failing, the programmer has no idea why it's failing. By my definition, if they don't know why it's failing, they could not have known how it works in the first place. If they don't know how it works, then they have no idea what they wrote in the first place. Just throwing crap at the wall.
When my code fails, I nearly always have an immediate theory as to why it's failing and most of the time my theory is perfectly correct or nearly so. Heck. Even when helping other people with their projects and figuring out why their code is not working correctly, my intuited guesses are better than their educated and informed theories. When I make these kinds of mistakes myself, I will fester on my mistake for weeks. I am my own worst critic and I hold myself to a high standard.
It's more rewarding to produce a system that is difficult to break. Breaking a sandcastle is fun, but building a sand castle that can't be broken is even more fun. When it is eventually broken, you learn from your mistake in lack of creativity.
I hate it when programmers blame the libraries they use. Never write a line of code or use a library unless you know what it's doing and how it's doing it. When I use a library, I'm hitting up github and reading their code.
Net neutrality is about the delivery mechanism being fair for all sources for a class of data. And "fair" being defined as something like "not interfering" or "not preferring". Peering with someone is not "preferring" because it does not manipulate the transmission in any way, it only brings the source "closer".
All of the big companies have found out how difficult it is to hire top talent by offering truck loads of money. If I'm in the top 0.1% of where I live and I love where I live, why would I care to get more money? There's actually been studies about offering more money and results. Offering too little is detrimental, but offering too much can be nearly as bad. Most people just want to be paid enough to do what they want to do an feel secure in their financial situation.
In my limited experience, skill and ability are inversely related. Highly skilled people are people who spent a lot of time doing a few things. They don't understand anything outside of the few things they do. In the end, they have this great skill of hammering nails, but they can't tell the difference between a screw an a nail.
Quite a bit of research is starting to show that in tech, masters are rarely very skillful and experienced workers are rarely masterful. Mastery comes from a deep understanding, which can be acquired independently of knowledge or experience. In many other professions, mastery and experience are highly correlated, but not so with highly cognitive and creative professions. A surgeon needs to physically experience the difficulty of suturing with blood soaked hands, but a programmer does not need to experience not validating input from an un-trusted source to know something bad will eventually happen. In the programmer's case, most of their issues can be 100% reasoned about, no experience required.
In problem domains where reasoning is almost the only thing you need to do to "learn", people who learn by thinking instead of doing can learn hundreds or thousands of times faster. The saying is it takes 10,000 hours to become a master. That is a generalization. It's 1,000-100,000 hours depending on the person and for domains where practice matters. Someone who becomes a master in 1,000 hours does so via deliberate practice, while the person who takes 100,000 hours via accidental practice. For domains where practice has virtually zero correlation, like programming, it becomes even more pronounced with more like 100-1,000,000 depending on the person. One person becomes a master in 2 weeks and another takes 350 years.
Not to mention that tech skills have a half-life of around 3-5 years. If it takes you more than 1 year to master something, it'll be useless by the time you master it. Domain transference is an ability of Abstract Reasoning. If you have strong Abstract Reasoning, you don't learn skills, you learn patterns. When you need to learn a new skill, you don't need to learn an entirely new skill, you abstract it into its fundamental patterns and transfer skill/knowledge from another domain.
I haven't had to learn a new tech skill in decades. Most of my tech skills I learned before I even owned a computer. The only skills I have remaining are non-tech related, like cleaner code, working as a team, communications, documentation, and other soft skills. Tech skills have been mostly unchanging. Just learning new tools for the same old problems that were solved back in the '60s.
I should add my family is a fun-loving group of witty people who love a good practical joke and a nice relaxing beer or two while hanging out with friends. When it comes to work, keep up or get out of the way. We do go well out of our way to help people and we enjoy being part of a team when the opportunity arises. We hate jerks and we loath drama.
We do have our black sheep. I was told my one uncle used to be normal, but then got into drugs for a decade. No one really knew where he was most of the time because he was coked out of his skull. Even after he got done with his drug spree because he was messed up in the head, he tested so well, he got offered several free rides through a few Ivy League Universities. My grandma had to turn them down because she had to take care of him. He may have lost most of his grip on reality, but he was still freaking smart, which was a scary mix when he started talking about the devil taking over the world. He kept family reunions lively.
My mom only has a GDE, and is a 55 year old secretary. She only recently got a computer and uses it maybe a few times a month. I can talk to her about optimizing distributed computing better than the CS graduates that I have to work with. Even my brother was a sophomore in college when he was leading 6 different graduate research teams at the same time. They had to take him off the teams and place him on his own because he was doing all of the work. The graduate students could not keep up. During his junior year in college, he had research Universities that dealt in a range of specializations from Super Computing to AI research, vying for him for a paid internship. In the end he took the super computing internship. Got paid $10k for the summer with room and board for free. Plus exclusive use of the entire super computer, normally reserved for PHDs. during his senior year, he did a one man $3mil datacenter project, also got to work with the FBI in enhancing their security, and taught several classes at the Uni.
Pretty much everyone in my family can talk to each other as peers, even if we're in different specialties. We have difficulty working as peers with most people, but we rarely need to seeing we have few peers. No one has called me an "arrogant asshole" because I enjoy being around happy people and see no reason to piss people off. My entire life while in school I was considered stupid until they found out I was gifted and just think differently. I did not like being judged, so I don't like judging others.
Don't go thinking everybody's life experience is the same as yours. If *you* hit the jackpot and got a great job straight out of college with no experience, don't assume experience is therefore worthless for anyone and everyone.
That wasn't what I was going for. By definition, Abstract Reasoning is the ability to solve problems without experience or knowledge. If the discussion is about domains like Computer Science or general programming where strong Abstract Reason is considered an absolute must, then experience and knowledge are by definition moot. Part of the definition of Abstract Reasoning is the ability to synthesize knowledge. If you need knowledge, make it, don't learn it. Of course this only applies to the technical aspects.
If *you* hit the jackpot
Then nearly everyone in my blood family hit the jackpot, on their own, with no help from their parents or the rest of the family.
Experience is nearly worthless. My brother was making $100k/year working 20 hours per week while in college, placing him in the top 0.1% of the local household income. Then he graduated, immediately with a job for $120k/year plus great benefits. In less than a month, he was promoted to the head of his department. He had the job lined up for over a year. They didn't care how long it took my brother to graduate, they were saving him a position.
They are not paying top dollar for my brother's experience, but his ability to solve difficult problems. He was lucky that he got a local job. My professors told me they knew people in high places at some of the biggest companies and were willing to get me a good job, but I turned them down because I would have to leave the state and be away from family.
Experience teaches you how to not make the same mistakes, but it does not teach you how to not make mistakes in the first place. Experience and knowledge are the worst ways of learning problem solving.
The poor people I've seen have iPhones, just a model or two older. I can get an iPhone 5 with an unlimited dataplan for $20/m cheaper than a basic cell phone.
Correction, lucky LPB
After looking more into it, it seems that rule of thumb is great for most situations, and probably especially most videos. For relatively static images, the effective acuity of the eye goes up as it re-scans the image. Even at 294 pixels per inch at arm's length, each pixel is going to stimulate around a hundred nerves. That means the underlying raw optical sensory of the eye is still 100x higher than the rule of thumb, we just don't get to directly perceive it for many reasons related to pre/post perceptual filter and pruning of visual data.
I'm getting the feeling of apples to oranges, similar to the "FPS" debate about how fast the eyes can see. There are several dimensions in which the eye can be measured, and normalizing it down to a single dimension of "frames per second" does not properly capture the complex situation.
I downloaded a PDF recently onto my S7 where the font size was quite small. So small that the capital "T" is about the size of the dot in "i" on my 23" 1080p monitor at font size 8. I was able to read this document without zoom at slightly over 12" away. Capital letters were only slightly larger than a few pixels on my monitor, yet I could read them with only minor strain and slightly reduced speed.
I was reading an interview with a VR specialist that said VR won't be able to do photorealistic rendering until somewhere between 8Mp and 16Mp per eye. While 720p is technically good enough from the rule of thumb, the rule of thumb breaks down on certain extreme edge cases. If we want VR to look like the real world, we're going to need between 16 and 32 Megapixels, at 300-1000 frames per second and about 16 bits of color depth for contrast.
Sounds like big numbers until you put them into perspective. Infrastructure and transit bandwidth is about 1%-3% of an established ISP's overall operating expenses. Network equipment and running the fiber constitutes only about 20% of the up-front costs. About 95% of the operating costs of an ISP are Marketing and customer support. One support truck-roll will cost an ISP nearly all of their profits for the customer for a year. Spending 15 minutes on the phone with a customer pays for nearly an entire year's worth of transit bandwidth and network upgrades.
The biggest hurdle to becoming and ISP is not the money, it's the red tape. Even if you get the venture capital, you will be held up by the local monopoly by denying you access or keeping you in court. ISPs are not federally recognized to have access to right-of-ways, only Cable and Telcoms, of which ISPs are neither. The easiest way to gain access to right-of-ways is to become a Cable company or Telcom, but then you immediately get heavily regulated and much of your profits go out the window.
It's nearly impossible to find ISPs that are only ISPs for these reasons. Our fastest internet is in the sticks where no big ISPs want to go. I can get a dedicated 500/500 fiber internet at a farm in the middle of no where for $100/m from a local private ISP with zero government subsidies. Or I can go to the local city and get 60/4 cable from Charter for $100/m, and rarely get 60 during peak hours and fight through the peak time lag and loss.
And boy do I mean dedicated. You will get your full speed 24/7 to nearly the entire world while having less than 1ms of jitter and less than 0.00001% average loss. I can go weeks without losing a single ICMP packet to Germany from Midwest USA or more than a brief 4ms flutter. My min ping is equal to my avg ping and my std-dev is less than 0.1ms from the average. Even under DDOS testing, my ping is never more than 20ms greater than the norm, but my loss goes sky high.
Even 384kbps mp3 sounds like crap. May not be able to tell the difference between 192 and 128, but I can easily tell the difference between lossless and mp3. Mp3 compression can cause my physical pain in my ears. I have no idea how this does not bother other people. I will get what feels and sounds and feels like swimmer's ear after a few minutes then my ears will start ringing and give me physical pain like an ear infection after a few more minutes. Ogg does not do this and of course not flac.
Most DJs I know use flac, but every so often an mp3 makes its way in and I either have to leave the room for a bit or ask the DJ to change the music. I typically just leave. If it's a DJ I'm paying for, I just tell them ahead of time not to play any mp3s.
Eyes have have a definition of about 6 megapixels in the center 5%. Perception filters out most of the details because they're not needed, but any unexpected visual data will stand out, like compression artifacts. If all of your vision was as good as the center, we'd need about 300Mp-400Mp per eye. As it stands, 75% of our visual acuity is in 5% of our vision. A non-uniform distribution with an extremely high standard deviation makes taking about averages meaningless.
DevOps is a cultural mindset, not a position.The job of an admin is to run a server. If you go to the cloud, with proper devops, your admins are pretty much out of a job.
QoS is only effective when there is not enough bandwidth. When pushed against the limitations of technology, QoS is expensive. Would you rather have 120Gb/s with QoS or 600Gb/s with no QoS. These are the choices that must be made.
That's not the legal definition, just the ideal definition.
A la carte works best when you don't want every channel under the Sun. If you want it all, sticking with your cable provider is probably your best bet. The point is most people don't want everything. At least the bills are more predictable. For a while on Charter, my bill was going up $2-$3 every month or two with no explanation while I was in contract. After a year of it, I had my wife call them up and start a fire storm. My wife can be down right evil on the phone, but only because she knows you have the play the game to get anywhere.
I just don't want any billing shenanigans. If I can get that while paying within 10%-20% for exactly the channels I want, that's a win.
The average density of the universe is roughly one atom for every four cubic meters.
This is why it's too expensive for ISPs to roll out fiber in the USA compared to Japan or Korea.
Quantum computers change symmetric key encryption from O(2^n) to O(2^(n/2)), saying nothing about the time take for each operation. But quantum computer can reverse certain types of private key crypto in the same number of operations as it takes to encrypt it, rendering those cryptos completely useless. Ignoring that currently adding more bits to a quantum computer is incredibly hard. Certain aspects of these algorithms could be scaled up to to where the quantum computer could not hold all of the state. RSA scales very poorly in size and people hate increasing the key size.
There is working being done to create an asymmetric crypto that has no quantum weaknesses. I think Lattice encryption is one such.