Sorry dude, but whoever told you that lied to you. In 2001, the USPS pension fund had $0 in it, and they'd promised to pay out $32 billion. You can see it in their annual report:
Congress mandated that they start catching up, first by reducing the outflow in 2003-2005, then by actually setting money aside starting in 2006. What was pretty extreme about the mandate is that Congress gave them only ten years to get caught up. That's roughly the same as asking someone to pay off their mortgage in 10 years - it can be done, but it sure isn't easy. That's what all the shouting was about - until the tinfoil-hat crowd started making up complete BS like what you spouted. Of course, there was a reason for the relatively short timeline - it's entirely possible that in the next ten years marketers will largely switch to email and stop using the postal service much. If that happens, USPS will not be bringing in revenue to cover their retirement plans. So the timeline is short, getting caught up is painful, but it would be unfair to their workers to rely on Penny Saver to still be pumping billions into the USPS 20 year from now.
Someone lied to you. When a private business tries to pull the shit USPS did with their retirement plan, someone goes to jail.
What's being required of USPS is that they do what everyone else does - if you promise today's employees that you'll pay them for today's work, but pay them later, you have to actually set aside money to make that payment. What they were doing is telling people working today that they'd get billions of dollars in pay 10, 20, or 30 years from now when they retire. They weren't putting any money aside to fund those promises, though.
When a private employer makes a promise like that, not only do they put money aside, they almost always hand it over to a third party trustee so that the company couldn't even spend the money if they wanted to.
Here's the thing about long term into the future. USPS was decades behind, they were having trouble paying people who worked 30 years ago with money USPS just got last year. Basically, they were 30 years in hole. Congress said that they needed to a) stop getting further behind and b) have a long-term plan to eventually get caught up, so that 30 years from now they'll be paying for work as it's done, not promising to pay for it years later.
In 1962, he sat down to write one book of 12 chapters. More than 50 years later, he hasn't finished the project, which has scope-creeper to seven books. In the process, he created several new languages for use in his book project, and became famous.
Yes, I think Knuth may simultaneously be the world's best programmer and the world's worst project manager.
The link you posted says it's fallacious to point out that SOMETHING is worse, that the thing under discussion isn't the WORST choice. That's a mistake because we should be looking for the BEST choice, not merely avoiding the worst one.
GP is arguing that Flash is the BEST choice in some scenarios, that ALL options are worse. That's fundamentally different from arguing that SOME options are worse. GP's argument is perfectly logical. Whether or not all of the alternatives actually ARE worse is another question, of course.
I CAN program in Flash, but I've never CHOSEN Flash when the choice was mine.
It can be very hard to let go, to trust the new person with your baby. I think it's generally true that "Whoever can be trusted with very little can also be trusted with much, and whoever is dishonest with very little will also be dishonest with much", so I didn't hand over everything on day one, but I tried not to delay unnecessarily once someone has proven themselves. Maybe give the new person full reign on a NEW project or system, something that won't wreck the company if it blows up. Similarly, maybe a just non-critical systems for a few weeks.
The other thing I tried to do that helped both the new person do the job and helped me feel better was to frequently communicate about how often the new person is asking questions. I try to guide them to ask questions when they need to (don't guess when you don't know), and also trust their knowledge when they do know. So I tell new people I plan to get questions from them. Even if they are an expert in their field, they aren't an expert on OUR systems. Also, experts talk to other experts. I'm part of several groups, standards bodies and FOSS development efforts, where highly competent people discuss ideas. So they should feel free to ask questions when needed - if they didn't any questions in their first week that would signify a problem. Conversely, I had one employee who would keep asking the same questions. She would check with me even when she knew what needed to be done. I had to encourage her to trust her own knowledge.
Knowing that my people ask intelligent questions about the areas they are responsible for, I know what I can trust them with. In those areas where they ask "dumb" questions and really don't know the answer, I know I shouldn't give them that specific responsibility outside of a learning projects.
> If you want multiple consoles, it's not like every Linux/BSD/OS X/Solaris distro doesn't have a button right there for you to open a freaking bazillion of them!!!
While you can use screen for that, that's not the most common/useful use case these days. One very handy use case is a session from a mobile device. A 4G connection is often briefly interrupted. With screen, an interruption in your connection doesn't reset your session. So you can continue to edit a file as your devices is handed off to different towers, and not lose anything when you're between towers. Automatic reconnect in your SSH client is handy in this case too.
Other variants of this include persistent sessions between devices - you can start working on your desktop, then pick right up where you left off on your laptop.
You can also have multiple task-based sessions going remotely. So you might have a "handle email" screen set up and a "program web site" screen set up. You can SSH in and choose either environment.
>> You pay Verizon to connect you to the internet. > Actually: I pay Verizon for a connection to NetFlix.
Verizon sells Netflix service? Where can I find that service plan on their web site? I only see internet service.
> I started in the 1980s, so I win
If you've been running networks for more than a month, you know Cogent IS the cheapest, crappiest, upstream you can buy. Netflix discovered WHY Cogent IS the cheapest - that Cogent doesn't consistently provide top quality. You and I have known that for 20 years.
When Netflix pays their hosting bill, like every other web site does, the benefit to consumers is that they can watch Netflix. Netflix was trying to be the only web site in the world who doesn't pay their bandwidth bill, without having their service cut off for nonpayment.
Initially, Netflix tried what a lot of new webmasters try - buying bandwidth from the cheapest, crappiest provider there us, Cogent. Cogent IS cheap. They are cheap because they don't consistently deliver the bandwidth they promise, and what they promise is low quality (high jitter, medium-high latency). Like most people running web sites, Netflix has discovered that if you want high quality, reliable bandwidth, you need to get it from someone other than Cogent.
Now Netflix is doing what we, and most successful web sites, have always done - buy bandwidth from other companies as well Cogent. You set up your routers to use Cogent when their connection to the destination happens to be good at the moment. If Cogent isn't fast to that destination at the moment, you route the packets via a better, more expensive provider.
Check the IETF standards discussions for IPv6 and HTTP back in the 1990s. You might see that I've been groking (and designing) this internet thing for a while.
Where you're being tricked is that you don't realize Netflix is basically just asking for free web hosting. Free hosting for the highest bandwidth site in the world. Netflix is NOT a peer of Verizon, so their attempt to call their upstream connection "peering" is misleading.
You pay Verizon to connect YOU to the internet. Netflix wants to pay nobody to connect them to the internet. They wanted Verizon to provide them with free bandwidth by providing multi-gigabit connections for free. Sorry Netflix, if you want gigabits of upstream bandwidth you have to pay for it, just like the rest of us.
This is about Netflix switching which provider they use, and Netflix paying their provider for bandwidth. It's exactly like Amazon paying their own bandwidth bills, which they do.
Netflix decided to sell a service which requires a lot of bandwidth - HD video. Now they are crying about paying their bill. I set up the servers for one of the first video sites on the internet back in the 1990s. All that bandwidth was expensive. We paid our bandwidth bill. Fortunately, our calculations were correct - though we paid alot more for bandwidth than the picture sites did, we also got a lot more users, which brought more revenue.
The fact that Slashdot is treating this as news seems to be confusing some people. Netflix just bought bandwidth from Verizon service, just like a million other people do. You pay for your connection, Netflix pays for theirs.
As the summary mentions, but apparently not clearly enough, this has nothing whatsoever to do with net neutrality. Netflix was getting a connection from Cogent (like I do). Now they are getting a connection from Verizon.
Since they use a lot of bandwidth to alot of places, they buy connections from several ISPs, again just like I do, and everyone else who runs popular web sites. That's how it's done and how it's always been done. The only thing new is that Netflix is whining about paying their bills.
I'm sorry you got fucked. To avoid putting yourself in that situation again, you might want to do two things. First, recognise that vendor lock-in is a risk to the enterprise, and that risk has an accountable cost. When you choose to be locked into TWO vendors, the software vendor AND a supported version of IE, your risk is the multiple of two components.
Secondly, when you find yourself in a situation where such a risk seems unavoidable, broaden your perspective to look at the business processes that create that context. Perhaps there is no acceptable software that meets the defined requirements. In that case, you can take another look at the requirements from a broader enterprise perspective.
As you may know, I've been running businesses for 25 years, and we've NEVER put ourselves in the position of sole-vendor risk like that. It takes forethought, but it absolutely is possible to avoid that situation.
If your employer did that before you arrived, or over your strong objections, then you followed my advice - you didn't buy garbage. Unfortunately someone else did.
However, I've dealt with a few different businesses and can't think of such a situation where all three leading solutions are ActiveX / IE only. I can think of one where for the GUI, you had to choose between ActiveX, Java, or a local client. A network CLI was also available. I'm curious what case you have in mind?
If I ddid run into a theoretical situation where a critical piece of software would rely on ActiveX, and therefore put the enterprise at the mercy of changing IE versions, I'd look at the broader picture and evaluate the business processes that are setting up that risk.
Your link supports none of your claims. Your claims are:
SEGS used 90% solar, 10% natural gas SEGS provided primary power at night SEGS "was eventually decommissioned"
The wiki article mentions none of those things. The references at the bottom of the article do, however, address those, and they say you are mistaken on all three points.
Other documents in the references section show that the design is that they need to run it on gas 25% of the time. Traditional power plants are used to provide power for those customers when it's not sunny.
The 18 Gw power facility was certainly one of the two primary reasons that the dam was built.
> Most cars have a stereo. Generally we don't attribute RTA deaths to FM radio though.
I include accidents while running errands in "auto fatalities", even if the car is also used to get to and from work. Even if, on the insurance paperwork, the owner said the car is "primarily" used to get to work and back.
> Or hydro, or geothermal, or wind, or lots of other things.
You couldn't think of the "lots of other things" because you actually pretty well covered it - geothermal and wind are the two "alternative" sources other than solar*. Each of the three, especially wind and geothermal, can make an important contribution. There is a full exposition in the paper linked above, but each of the three can provide for 3%-5% of our energy needs. Combined, they can reduce fossil fuel use by 9.6%, which is quite significant.
* Maybe one day biomass will become significant, but it hasn't yet. There's ethanol in gas, but that actually reduces fuel economy enough that the net energy production is roughly zero after including the energy to produce the ethanol. Ethanol is a serviceable oxygen donor in gasoline to replace MTBE, but not currently a net energy producer outside of Brazil.
Sure, I already have some radioactive stuff glowing under my bed, and I put that there on purpose. I also put something that is radioactive, emitting alpha particles, in my baby's room. I know it's safe because the smoke detector is more than 2cm from the kid, and the radioactive part has a bit of plastic over the front.
Of course, if one were concerned, you could measure the levels to make sure it is just plutonium - that there is no uranium mixed in from another reactor or something.
Absolutely. If people gave a shit about the environment they'd use and promote solar thermal, which actually works. It can be as simple as an exterior water tank painted black, so it preheats the water before it enters the natural gas water heater.
Becauseit's simple and actually works, it's less fun to talk about than perpetual motion, aliens, Nikola Tesla, and magical solar panels that work at night.
Your smoking something. Unless of course you're talking about a "power plant" that only operated from 11AM to 1PM. In some places, solar IS a nice supplement during those hours.
> With nuclear the probability for a serious accident is non zero, and the stakes are much higher than coal for instance.
Higher than coal, you day. Let's compare coal. There have been about 500,000 casualties from black lung, and the indirect, environmental damage is incalculable. Compare to about 200 people from nuclear power. Right now there are 600 nuclear power plants operating, and we've had nuclear power for about 50 years, so we have a good basis of comparison. Nuclear is, by the numbers, at least 2,000 times safer than coal.
Are there risks with everything? Of course! Nuclear has risks. Modern designs don't have the catastrophic risks of 1950s designs, but of course there are risks. We need power, though, so we need to look at the safEST options. Nothing is perfectly safe - taking a shower kills a lot of people. Of the available options, nuclear is safER than anything else that can produce the gigawatts of stable, reliable power that we need.
Slashdot Mirror
Sorry dude, but whoever told you that lied to you. In 2001, the USPS pension fund had $0 in it, and they'd promised to pay out $32 billion. You can see it in their annual report:
http://about.usps.com/who-we-a...
Click "balance sheets" and it's at the bottom of page1.
Congress mandated that they start catching up, first by reducing the outflow in 2003-2005, then by actually setting money aside starting in 2006. What was pretty extreme about the mandate is that Congress gave them only ten years to get caught up. That's roughly the same as asking someone to pay off their mortgage in 10 years - it can be done, but it sure isn't easy. That's what all the shouting was about - until the tinfoil-hat crowd started making up complete BS like what you spouted. Of course, there was a reason for the relatively short timeline - it's entirely possible that in the next ten years marketers will largely switch to email and stop using the postal service much. If that happens, USPS will not be bringing in revenue to cover their retirement plans. So the timeline is short, getting caught up is painful, but it would be unfair to their workers to rely on Penny Saver to still be pumping billions into the USPS 20 year from now.
They've done a pretty good job catching up, as you can see in this 2014 GAO report:
http://www.gao.gov/assets/670/...
Someone lied to you. When a private business tries to pull the shit USPS did with their retirement plan, someone goes to jail.
What's being required of USPS is that they do what everyone else does - if you promise today's employees that you'll pay them for today's work, but pay them later, you have to actually set aside money to make that payment. What they were doing is telling people working today that they'd get billions of dollars in pay 10, 20, or 30 years from now when they retire. They weren't putting any money aside to fund those promises, though.
When a private employer makes a promise like that, not only do they put money aside, they almost always hand it over to a third party trustee so that the company couldn't even spend the money if they wanted to.
Here's the thing about long term into the future. USPS was decades behind, they were having trouble paying people who worked 30 years ago with money USPS just got last year. Basically, they were 30 years in hole. Congress said that they needed to a) stop getting further behind and b) have a long-term plan to eventually get caught up, so that 30 years from now they'll be paying for work as it's done, not promising to pay for it years later.
In 1962, he sat down to write one book of 12 chapters. More than 50 years later, he hasn't finished the project, which has scope-creeper to seven books. In the process, he created several new languages for use in his book project, and became famous.
Yes, I think Knuth may simultaneously be the world's best programmer and the world's worst project manager.
Starting in April 2012, most new systems had Linux pre-installed. Not coincidentally, that same year most new systems were pocket sized.
The link you posted says it's fallacious to point out that SOMETHING is worse, that the thing under discussion isn't the WORST choice. That's a mistake because we should be looking for the BEST choice, not merely avoiding the worst one.
GP is arguing that Flash is the BEST choice in some scenarios, that ALL options are worse. That's fundamentally different from arguing that SOME options are worse. GP's argument is perfectly logical. Whether or not all of the alternatives actually ARE worse is another question, of course.
I CAN program in Flash, but I've never CHOSEN Flash when the choice was mine.
It can be very hard to let go, to trust the new person with your baby. I think it's generally true that "Whoever can be trusted with very little can also be trusted with much, and whoever is dishonest with very little will also be dishonest with much", so I didn't hand over everything on day one, but I tried not to delay unnecessarily once someone has proven themselves. Maybe give the new person full reign on a NEW project or system, something that won't wreck the company if it blows up. Similarly, maybe a just non-critical systems for a few weeks.
The other thing I tried to do that helped both the new person do the job and helped me feel better was to frequently communicate about how often the new person is asking questions. I try to guide them to ask questions when they need to (don't guess when you don't know), and also trust their knowledge when they do know. So I tell new people I plan to get questions from them. Even if they are an expert in their field, they aren't an expert on OUR systems. Also, experts talk to other experts. I'm part of several groups, standards bodies and FOSS development efforts, where highly competent people discuss ideas. So they should feel free to ask questions when needed - if they didn't any questions in their first week that would signify a problem. Conversely, I had one employee who would keep asking the same questions. She would check with me even when she knew what needed to be done. I had to encourage her to trust her own knowledge.
Knowing that my people ask intelligent questions about the areas they are responsible for, I know what I can trust them with. In those areas where they ask "dumb" questions and really don't know the answer, I know I shouldn't give them that specific responsibility outside of a learning projects.
> If you want multiple consoles, it's not like every Linux/BSD/OS X/Solaris distro doesn't have a button right there for you to open a freaking bazillion of them!!!
While you can use screen for that, that's not the most common/useful use case these days. One very handy use case is a session from a mobile device. A 4G connection is often briefly interrupted. With screen, an interruption in your connection doesn't reset your session. So you can continue to edit a file as your devices is handed off to different towers, and not lose anything when you're between towers. Automatic reconnect in your SSH client is handy in this case too.
Other variants of this include persistent sessions between devices - you can start working on your desktop, then pick right up where you left off on your laptop.
You can also have multiple task-based sessions going remotely. So you might have a "handle email" screen set up and a "program web site" screen set up. You can SSH in and choose either environment.
http://tech.slashdot.org/comme...
Knuth can write a complete OS in assembly language while sleeping. 99.9999% of programmers aren't Knuth.
>> You pay Verizon to connect you to the internet.
> Actually: I pay Verizon for a connection to NetFlix.
Verizon sells Netflix service? Where can I find that service plan on their web site? I only see internet service.
> I started in the 1980s, so I win
If you've been running networks for more than a month, you know Cogent IS the cheapest, crappiest, upstream you can buy. Netflix discovered WHY Cogent IS the cheapest - that Cogent doesn't consistently provide top quality. You and I have known that for 20 years.
When Netflix pays their hosting bill, like every other web site does, the benefit to consumers is that they can watch Netflix. Netflix was trying to be the only web site in the world who doesn't pay their bandwidth bill, without having their service cut off for nonpayment.
Initially, Netflix tried what a lot of new webmasters try - buying bandwidth from the cheapest, crappiest provider there us, Cogent. Cogent IS cheap. They are cheap because they don't consistently deliver the bandwidth they promise, and what they promise is low quality (high jitter, medium-high latency). Like most people running web sites, Netflix has discovered that if you want high quality, reliable bandwidth, you need to get it from someone other than Cogent.
Now Netflix is doing what we, and most successful web sites, have always done - buy bandwidth from other companies as well Cogent. You set up your routers to use Cogent when their connection to the destination happens to be good at the moment. If Cogent isn't fast to that destination at the moment, you route the packets via a better, more expensive provider.
Check the IETF standards discussions for IPv6 and HTTP back in the 1990s. You might see that I've been groking (and designing) this internet thing for a while.
Where you're being tricked is that you don't realize Netflix is basically just asking for free web hosting. Free hosting for the highest bandwidth site in the world. Netflix is NOT a peer of Verizon, so their attempt to call their upstream connection "peering" is misleading.
You pay Verizon to connect YOU to the internet. Netflix wants to pay nobody to connect them to the internet. They wanted Verizon to provide them with free bandwidth by providing multi-gigabit connections for free. Sorry Netflix, if you want gigabits of upstream bandwidth you have to pay for it, just like the rest of us.
This is about Netflix switching which provider they use, and Netflix paying their provider for bandwidth. It's exactly like Amazon paying their own bandwidth bills, which they do.
Netflix decided to sell a service which requires a lot of bandwidth - HD video. Now they are crying about paying their bill. I set up the servers for one of the first video sites on the internet back in the 1990s. All that bandwidth was expensive. We paid our bandwidth bill. Fortunately, our calculations were correct - though we paid alot more for bandwidth than the picture sites did, we also got a lot more users, which brought more revenue.
I'm glad you asked.
https://www.dropbox.com/s/cbkc...
Pages 12-14 contain full list of references for the data. Earlier pages provide analysis and discussion of the data.
The fact that Slashdot is treating this as news seems to be confusing some people. Netflix just bought bandwidth from Verizon service, just like a million other people do. You pay for your connection, Netflix pays for theirs.
As the summary mentions, but apparently not clearly enough, this has nothing whatsoever to do with net neutrality. Netflix was getting a connection from Cogent (like I do). Now they are getting a connection from Verizon.
Since they use a lot of bandwidth to alot of places, they buy connections from several ISPs, again just like I do, and everyone else who runs popular web sites. That's how it's done and how it's always been done. The only thing new is that Netflix is whining about paying their bills.
I'm sorry you got fucked. To avoid putting yourself in that situation again, you might want to do two things. First, recognise that vendor lock-in is a risk to the enterprise, and that risk has an accountable cost. When you choose to be locked into TWO vendors, the software vendor AND a supported version of IE, your risk is the multiple of two components.
Secondly, when you find yourself in a situation where such a risk seems unavoidable, broaden your perspective to look at the business processes that create that context. Perhaps there is no acceptable software that meets the defined requirements. In that case, you can take another look at the requirements from a broader enterprise perspective.
As you may know, I've been running businesses for 25 years, and we've NEVER put ourselves in the position of sole-vendor risk like that. It takes forethought, but it absolutely is possible to avoid that situation.
> When your employer chose to adopt one
If your employer did that before you arrived, or over your strong objections, then you followed my advice - you didn't buy garbage. Unfortunately someone else did.
However, I've dealt with a few different businesses and can't think of such a situation where all three leading solutions are ActiveX / IE only. I can think of one where for the GUI, you had to choose between ActiveX, Java, or a local client. A network CLI was also available. I'm curious what case you have in mind?
If I ddid run into a theoretical situation where a critical piece of software would rely on ActiveX, and therefore put the enterprise at the mercy of changing IE versions, I'd look at the broader picture and evaluate the business processes that are setting up that risk.
a) Don't buy garbage, stuff that works only in a specific version of a specific browser.
b) 90%+ plus, you can just set the user agent header in Seamonkey, Firefox, or Chrome to SAY it's IE and things work just fine.
Your link supports none of your claims. Your claims are:
SEGS used 90% solar, 10% natural gas
SEGS provided primary power at night
SEGS "was eventually decommissioned"
The wiki article mentions none of those things. The references at the bottom of the article do, however, address those, and they say you are mistaken on all three points.
Nextera Energy say they are still operating the SEGS plants:
http://www.nexteraenergy.com/c...
Other documents in the references section show that the design is that they need to run it on gas 25% of the time. Traditional power plants are used to provide power for those customers when it's not sunny.
The 18 Gw power facility was certainly one of the two primary reasons that the dam was built.
> Most cars have a stereo. Generally we don't attribute RTA deaths to FM radio though.
I include accidents while running errands in "auto fatalities", even if the car is also used to get to and from work.
Even if, on the insurance paperwork, the owner said the car is "primarily" used to get to work and back.
I'm glad you asked. Pages 12-14 contain full list of references.
https://www.dropbox.com/s/cbkc...
> Or hydro, or geothermal, or wind, or lots of other things.
You couldn't think of the "lots of other things" because you actually pretty well covered it - geothermal and wind are the two "alternative" sources other than solar*. Each of the three, especially wind and geothermal, can make an important contribution. There is a full exposition in the paper linked above, but each of the three can provide for 3%-5% of our energy needs. Combined, they can reduce fossil fuel use by 9.6%, which is quite significant.
* Maybe one day biomass will become significant, but it hasn't yet. There's ethanol in gas, but that actually reduces fuel economy enough that the net energy production is roughly zero after including the energy to produce the ethanol. Ethanol is a serviceable oxygen donor in gasoline to replace MTBE, but not currently a net energy producer outside of Brazil.
Sure, I already have some radioactive stuff glowing under my bed, and I put that there on purpose. I also put something that is radioactive, emitting alpha particles, in my baby's room. I know it's safe because the smoke detector is more than 2cm from the kid, and the radioactive part has a bit of plastic over the front.
Of course, if one were concerned, you could measure the levels to make sure it is just plutonium - that there is no uranium mixed in from another reactor or something.
Absolutely. If people gave a shit about the environment they'd use and promote solar thermal, which actually works. It can be as simple as an exterior water tank painted black, so it preheats the water before it enters the natural gas water heater.
Becauseit's simple and actually works, it's less fun to talk about than perpetual motion, aliens, Nikola Tesla, and magical solar panels that work at night.
Your smoking something. Unless of course you're talking about a "power plant" that only operated from 11AM to 1PM. In some places, solar IS a nice supplement during those hours.
> With nuclear the probability for a serious accident is non zero, and the stakes are much higher than coal for instance.
Higher than coal, you day. Let's compare coal. There have been about 500,000 casualties from black lung, and the indirect, environmental damage is incalculable. Compare to about 200 people from nuclear power. Right now there are 600 nuclear power plants operating, and we've had nuclear power for about 50 years, so we have a good basis of comparison. Nuclear is, by the numbers, at least 2,000 times safer than coal.
Are there risks with everything? Of course! Nuclear has risks. Modern designs don't have the catastrophic risks of 1950s designs, but of course there are risks. We need power, though, so we need to look at the safEST options. Nothing is perfectly safe - taking a shower kills a lot of people. Of the available options, nuclear is safER than anything else that can produce the gigawatts of stable, reliable power that we need.