Seems what we need is fiber ONTs that can accept power from an in-line power line that is integrated into the fiber cable, then when power goes out, the ONT falls into a low power voice only mode.
Fiber optics do no degrade at an appreciable rate. There should be almost no maintenance for fiber other than people digging where they should not. With our fiber being buried 18 inches underground, parallel to 3.3kv underground power lines and natural gas, you'd be a complete idiot to randomly dig into the fiber as you'd probably kill yourself hitting one of the other lines.
Not to mention, the trunk lines are on public property, so why would a civilian be digging there? If you dig in your own property, you have no chance to affect anyone else's fiber because you'll just cut your own line.
Enjoy paying the utilities to come out and fix your crap. You must pay for the entire cost.
Yes, a back-hoe could take out the fiber, but they'd also take out the electricity and the natural gas in the process. BOOM. They make sure they get their crap marked and don't dig unless they're sure.
Once you get outside the city, then things are different as they use overhead power and no natural gas lines, so a backhoe could take out the country-folk's fiber.
In general, because of the whole city thing, construction companies are quite good at flagging their stuff and not cutting lines. What does happen a lot around here is when the power goes out, it takes out the phone lines because the power usually goes out because of a tree falling on a pole, which also takes out the phone.
Because of this, quite a few areas have buried POTS. I see no benefit at all using copper over fiber, except that now we need a UPS to keep our fiber ONT powered in order to make phone calls during a power outage. Not that it matters, because I don't own a normal phone.
Being able to write good software is a gift that few people possess. We should be paid well to do it
Most good coders love to code. Getting paid is great for putting food on the table and all that boring stuff, but I don't share the "I think I should always get paid for my code". Most would work for free/fun if they could. Good coders should not feel entitled, but they should at least be given the choice. GPL effective forces programmers to drink the Koolaid, or to not participate. Kinds of sucks. Be excluded or be forced to not give out your talent for free. GPL is NOT free, but pretty close.
Kind of funny that most big name projects that run on Linux are actually BSD licensed.
You just try feeding a teenage boy on less than $1,000 a year.
You'd have to starve them or feed them all carbs. It costs me about $150/month to eat semi-healthy, which is almost $2,500 per year, and I eat a fraction of what I ate as a teenager. I miss being able to eat anything I wanted and still remain 160lbs.
And if it doesn't now you not only have a worse query plan, but you also have a temp table to store that sub-query. Good job being a bad programmer. Don't write a sub-query unless you actually want a sub-query. MS added "WITH" for just this reason. If you want to just clean up your code, use WITH and not a sub-query.
While most of your experience may have been the sub-query getting optimized out, most of my experience of slow SQL has been sub-queries NOT getting optimized out. Even worse, most people in my experience who use sub-queries also tend to accidentally create a correlated query.
Not to mention that the query plan can change depending on table meta-data. The query plan is less likely to get done incorrectly if you just used an inner-join where you meant an inner-join.
I think a lot of people think of "telecommuting" as just SSHing in and doing work, then bouncing around emails or voice communications. With better Internet connections, HD video communications will help with the situation. A good investment for telecommuting, I would think, would be a strong upload connection, a high quality camera, and a good microphone. There is a psychological element to it.
I still have ads enabled for sites that I like, and I recently saw one for a datacenter looking for a programmer with quite some experience, and a fast reliable Internet connection to telecommute for a full time position.
Yeah, the "opportunity hypothesis" would apply more back 15+years ago, but now computers are everywhere. Programmers see computers as a way to augment their abilities by instructing something to do the work for them. When I realized a computer's potential, it wasn't from playing with it for a long time, it was from seeing someone use it for 5 minutes. It was instantly obvious to me, even at a young age. Prior to seeing a computer, I didn't even know what they were, I never heard of them, I blissfully watched TV all the time. I suddenly had something to be excited about.
Around 10, I heard about the Internet, that was the second most suddenly exciting moment of computers. I had no idea what was on the Internet or how to use it, but allowing computers to communicate must be freaking awesome.
I did spend large amounts of time playing video games, but I also spent a lot of time reading about them.
Nearly all of your examples were during the time that their tech was just taking off. Once a market is mature, it's hard to become another Microsoft. When competition is low or non-existent, all you need to do is find some money. In this case, their luck was timing on when they were born. Timing is everything.
There may be a lot of successful people out there, but not a lot of ultra-successful billionaires. You're more likely to be born into money than make your own.
The tools change, but the same problems just keep recirculating. But once you understand the problems, you just need to figure out how the tools address the problems, and typically newer "Better" tools are easier to use. Technology changes, but problems do not. If you think they're changing, you're not understanding the problems.
It's almost a requirement to be lucky in order to be in a position of power. Ability, not as much. With both ability and luck, you can become very powerful. May want to look at a history book.
They tend to solve problems slightly differently, so it creates for a wider range of ideas. Quite often I find myself getting a few programmers together, discussing the problem domain, then throwing ideas. Getting a good list of ideas and their pros and cons is important.
Similar for me. I was that below average student with no strengths. I did well in discussions and understood stuff, but I had a hard time doing arithmetic in my head or memorizing stuff. Around 8, I got to see a computer. It instantly clicked with me. I didn't have to do the math, I only had to understand it, the computer did all the work for me. After that, I started doing lots of reading on ASM, C, the kind of math video games have to work with, how CPUs work, how memory works, how HDs work, no chipsets work, how network work, what kind of problems there are like latency vs throughput.
A large portion of my time at work is designing and theorycrafting systems. Thinking about all of the choke-points and how to solve the problem in a good way. Always thinking about worst cases a given choke point could have and what requirements we could need to scale up the chokepoints if those occurred. etc etc
And my high school adviser said I shouldn't go to college because of my bad grades.
I would argue that computer tech is quite stable. From an abstract level, new tech just makes it easier to do old problems. The problem domain of computers have not changed much, just what becomes new the flavor of the next 5 years. Same problems, just handled slightly differently. Once you understand the problems, you're good for a long time.
A car analogy would be something like the old way to get from point a to b was to walk, but now you can take a bus. Yes, you need to learn how to read a bus schedule, but it is the same underlying problem. You may also learn that while buses are great for throughput, they're bad for latency, so you might use a car instead.
Most people I've met can't be good coders because of their mindset. They just want something to work, they don't care about how or why and they take no pride. Nothing like trying to tell a another programmer why their design isn't good because they don't understand how the hardware works. Why use a hashtable when you can use a linked list? They both get the job done. Who cares if the design will triple the number of round trips to the DB, who cares that the DB tables are flat and unnormalized, who cares that there are no sprocs. It works, right? That seems to be 80% of the people I've met.
The government could punish the local phone companies for not enforcing higher quality peering relations. Essentially indirectly block clear abusers by threat of being blacklisted from the national community if you allow scammers to use your network.
I forget which company, but there was a datacenter in some other country that got caught hosting the services of a popular spammer. It took a long while to trace them down, but they got found out. The datacenter got told to stop it or face getting their routes blocked by all Backbone providers. Force those datacenters to screen their customers a bit better. Similar thing for VOIP services.
The "47%" that people keep talking about are almost entirely retirees, adults going to full time college, or the disabled. The largest portion of the group are the retired people. Are you saying that people who have worked their whole like and are now retired, should not have a say in taxes?
Just because I agreed to a contract, doesn't mean that I like it or think it's fair. What if I'm indirectly forced into it because I have no other options. To be fair is to be just and and unfair compensation is unjust. There is not issue with making all the money you want, until it starts to hurt others. What if I was a billionaire and started buying up all of the farms, then I refused to let people buy food from me? My farms! Let the people starve.
Too much of a good thing can be bad. If it can't come to a natural balance on its own, then it needs to be forced into an artificial balance or it needs to be destroyed.
I tried Dropbox a few months back. It was quite slow because of the way they transferred the data in a blocking way. When trying to upload a 2GB file to their service, I would see my network activity jump to about 30mb/s in 1/2 a second, then drop down to 0, then 1/2sec later, start transferring the next block. My connection was effective used only 1/2 the time. Even worse, TCP didn't have enough time to ramp up, so it couldn't make full use of my connection.
No idea. I was under the impression that most NICs have just enough onboard memory to buffer potential bursts of data, but otherwise write to system memory via DMA and interrupt the CPU to notify it when the data is ready. 512MB sounds like a lot of buffer for just a 1gb NIC.
100mhz DDR3 isn't that much faster for latency than EDO 66mhz, but it does have more bits to charge up and takes a bit longer to find the correct bits. The external bus is a lot faster, but the internal processing speed is not.
Slashdot Mirror
Seems what we need is fiber ONTs that can accept power from an in-line power line that is integrated into the fiber cable, then when power goes out, the ONT falls into a low power voice only mode.
Fiber optics do no degrade at an appreciable rate. There should be almost no maintenance for fiber other than people digging where they should not. With our fiber being buried 18 inches underground, parallel to 3.3kv underground power lines and natural gas, you'd be a complete idiot to randomly dig into the fiber as you'd probably kill yourself hitting one of the other lines.
Not to mention, the trunk lines are on public property, so why would a civilian be digging there? If you dig in your own property, you have no chance to affect anyone else's fiber because you'll just cut your own line.
Enjoy paying the utilities to come out and fix your crap. You must pay for the entire cost.
Yes, a back-hoe could take out the fiber, but they'd also take out the electricity and the natural gas in the process. BOOM. They make sure they get their crap marked and don't dig unless they're sure.
Once you get outside the city, then things are different as they use overhead power and no natural gas lines, so a backhoe could take out the country-folk's fiber.
In general, because of the whole city thing, construction companies are quite good at flagging their stuff and not cutting lines. What does happen a lot around here is when the power goes out, it takes out the phone lines because the power usually goes out because of a tree falling on a pole, which also takes out the phone.
Because of this, quite a few areas have buried POTS. I see no benefit at all using copper over fiber, except that now we need a UPS to keep our fiber ONT powered in order to make phone calls during a power outage. Not that it matters, because I don't own a normal phone.
Being able to write good software is a gift that few people possess. We should be paid well to do it
Most good coders love to code. Getting paid is great for putting food on the table and all that boring stuff, but I don't share the "I think I should always get paid for my code". Most would work for free/fun if they could. Good coders should not feel entitled, but they should at least be given the choice. GPL effective forces programmers to drink the Koolaid, or to not participate. Kinds of sucks. Be excluded or be forced to not give out your talent for free. GPL is NOT free, but pretty close.
Kind of funny that most big name projects that run on Linux are actually BSD licensed.
You just try feeding a teenage boy on less than $1,000 a year.
You'd have to starve them or feed them all carbs. It costs me about $150/month to eat semi-healthy, which is almost $2,500 per year, and I eat a fraction of what I ate as a teenager. I miss being able to eat anything I wanted and still remain 160lbs.
Popular doesn't mean good, only popular.
And if it doesn't now you not only have a worse query plan, but you also have a temp table to store that sub-query. Good job being a bad programmer. Don't write a sub-query unless you actually want a sub-query. MS added "WITH" for just this reason. If you want to just clean up your code, use WITH and not a sub-query.
While most of your experience may have been the sub-query getting optimized out, most of my experience of slow SQL has been sub-queries NOT getting optimized out. Even worse, most people in my experience who use sub-queries also tend to accidentally create a correlated query.
Not to mention that the query plan can change depending on table meta-data. The query plan is less likely to get done incorrectly if you just used an inner-join where you meant an inner-join.
I think a lot of people think of "telecommuting" as just SSHing in and doing work, then bouncing around emails or voice communications. With better Internet connections, HD video communications will help with the situation. A good investment for telecommuting, I would think, would be a strong upload connection, a high quality camera, and a good microphone. There is a psychological element to it.
I still have ads enabled for sites that I like, and I recently saw one for a datacenter looking for a programmer with quite some experience, and a fast reliable Internet connection to telecommute for a full time position.
Yeah, the "opportunity hypothesis" would apply more back 15+years ago, but now computers are everywhere. Programmers see computers as a way to augment their abilities by instructing something to do the work for them. When I realized a computer's potential, it wasn't from playing with it for a long time, it was from seeing someone use it for 5 minutes. It was instantly obvious to me, even at a young age. Prior to seeing a computer, I didn't even know what they were, I never heard of them, I blissfully watched TV all the time. I suddenly had something to be excited about.
Around 10, I heard about the Internet, that was the second most suddenly exciting moment of computers. I had no idea what was on the Internet or how to use it, but allowing computers to communicate must be freaking awesome.
I did spend large amounts of time playing video games, but I also spent a lot of time reading about them.
Nearly all of your examples were during the time that their tech was just taking off. Once a market is mature, it's hard to become another Microsoft. When competition is low or non-existent, all you need to do is find some money. In this case, their luck was timing on when they were born. Timing is everything.
There may be a lot of successful people out there, but not a lot of ultra-successful billionaires. You're more likely to be born into money than make your own.
The tools change, but the same problems just keep recirculating. But once you understand the problems, you just need to figure out how the tools address the problems, and typically newer "Better" tools are easier to use. Technology changes, but problems do not. If you think they're changing, you're not understanding the problems.
It's almost a requirement to be lucky in order to be in a position of power. Ability, not as much. With both ability and luck, you can become very powerful. May want to look at a history book.
They tend to solve problems slightly differently, so it creates for a wider range of ideas. Quite often I find myself getting a few programmers together, discussing the problem domain, then throwing ideas. Getting a good list of ideas and their pros and cons is important.
Similar for me. I was that below average student with no strengths. I did well in discussions and understood stuff, but I had a hard time doing arithmetic in my head or memorizing stuff. Around 8, I got to see a computer. It instantly clicked with me. I didn't have to do the math, I only had to understand it, the computer did all the work for me. After that, I started doing lots of reading on ASM, C, the kind of math video games have to work with, how CPUs work, how memory works, how HDs work, no chipsets work, how network work, what kind of problems there are like latency vs throughput.
A large portion of my time at work is designing and theorycrafting systems. Thinking about all of the choke-points and how to solve the problem in a good way. Always thinking about worst cases a given choke point could have and what requirements we could need to scale up the chokepoints if those occurred. etc etc
And my high school adviser said I shouldn't go to college because of my bad grades.
I would argue that computer tech is quite stable. From an abstract level, new tech just makes it easier to do old problems. The problem domain of computers have not changed much, just what becomes new the flavor of the next 5 years. Same problems, just handled slightly differently. Once you understand the problems, you're good for a long time.
A car analogy would be something like the old way to get from point a to b was to walk, but now you can take a bus. Yes, you need to learn how to read a bus schedule, but it is the same underlying problem. You may also learn that while buses are great for throughput, they're bad for latency, so you might use a car instead.
Most people I've met can't be good coders because of their mindset. They just want something to work, they don't care about how or why and they take no pride. Nothing like trying to tell a another programmer why their design isn't good because they don't understand how the hardware works. Why use a hashtable when you can use a linked list? They both get the job done. Who cares if the design will triple the number of round trips to the DB, who cares that the DB tables are flat and unnormalized, who cares that there are no sprocs. It works, right? That seems to be 80% of the people I've met.
The government could punish the local phone companies for not enforcing higher quality peering relations. Essentially indirectly block clear abusers by threat of being blacklisted from the national community if you allow scammers to use your network.
I forget which company, but there was a datacenter in some other country that got caught hosting the services of a popular spammer. It took a long while to trace them down, but they got found out. The datacenter got told to stop it or face getting their routes blocked by all Backbone providers. Force those datacenters to screen their customers a bit better. Similar thing for VOIP services.
The "47%" that people keep talking about are almost entirely retirees, adults going to full time college, or the disabled. The largest portion of the group are the retired people. Are you saying that people who have worked their whole like and are now retired, should not have a say in taxes?
Your English is better than most! Really from Ukraine?
If people don't have benefit from it, then it's moot. You measure healthcare by how it affects people, in our case, we have some of the worst.
Just because I agreed to a contract, doesn't mean that I like it or think it's fair. What if I'm indirectly forced into it because I have no other options. To be fair is to be just and and unfair compensation is unjust. There is not issue with making all the money you want, until it starts to hurt others. What if I was a billionaire and started buying up all of the farms, then I refused to let people buy food from me? My farms! Let the people starve.
Too much of a good thing can be bad. If it can't come to a natural balance on its own, then it needs to be forced into an artificial balance or it needs to be destroyed.
I tried Dropbox a few months back. It was quite slow because of the way they transferred the data in a blocking way. When trying to upload a 2GB file to their service, I would see my network activity jump to about 30mb/s in 1/2 a second, then drop down to 0, then 1/2sec later, start transferring the next block. My connection was effective used only 1/2 the time. Even worse, TCP didn't have enough time to ramp up, so it couldn't make full use of my connection.
No idea. I was under the impression that most NICs have just enough onboard memory to buffer potential bursts of data, but otherwise write to system memory via DMA and interrupt the CPU to notify it when the data is ready. 512MB sounds like a lot of buffer for just a 1gb NIC.
Submarine simulator, get it right!
100mhz DDR3 isn't that much faster for latency than EDO 66mhz, but it does have more bits to charge up and takes a bit longer to find the correct bits. The external bus is a lot faster, but the internal processing speed is not.