The 787 program was started 2-3 years after the A380, and a significantly more difficult engineering task than the 380. Airbus had the same kinds of problems Boeing did in their supply chain. Airbus had to cancel their freighter version. Both companies have massive government subsidies that the other side's government proclaim are illegal.
Boeing was obsessed with the Sonic Cruiser before the 787, which is why they ended up behind the game compared to Airbus.
Personally, my favorite plane to fly will always be the upper deck of a 747, as it creates an intimate space where you forget about all the other people on the plane. The 380 is more like to 777's stacked on top of each other. It might make sense in some markets but it is hard to believe that it will be the long-term solution.
Ground temperature at the surface to a reasonable depth (not sure exactly what that is) is equal to the average annual temperature. The problem is doing horizontal pipes below the frost line vs. vertical bores spaced adequately apart.
We looked at doing a geothermal project for a limestone mine to be converted to a data center, but it wasn't practical as they filled it in to reduce flooding risk and chamber height. Best approach is heating/cooling the aquifer, but that has similar problems at a macro-scale.
The real issue is if you want to be in a technical role for life, or if you would like to transition to management. The initial change from technical to managerial positions means a sacrifice, but long-term there are significant benefits.
In my field, people below the line get paid overtime, and above the line are pure salary. Many people get 10-15% overtime, and only a 5-7% pay increase for crossing the line, thus taking a pay cut in the process. Within a year or two they usually make up the gap and then some, as they are eligible for higher bonuses instead.
If you think you are just going to become a worthless PHB and that has no interest for you, by all means pass though...
If you want to see computers saving hospitals money, get a physical at Bumrungrad Hospital in Bangkok. The system is multi-lingual, and you get results almost immediately. I don't know that a legacy hospital can pull off the same things, but it is truly amazing to see.
Compared to my physician's scribbles on pre-printed form letters, it really does a lot to make it look like the doctor knows what they are doing and communicate effectively.
Even if you are going co-lo, you are best off hiring a pro to help you. (Disclaimer: I am a pro.) The checklist approach is a good starting point, but far too often you will miss site-specific issues.
How old are the UPS batteries? Where are the preventative maintenance logs? What happens if the facility is expanded? Are there any single points of failure you need to understand? Is it 6N/5 or N+1? Are there lease terms that you need to be careful with?
We have done big deals with large customers, and despite having a good internal team, good legal experts, and it is always amazing what little things come up that are really critical to addressing properly to protect liability.
Purchasing a facility is often easier; you either know the facility is a write-off, has some level of residual value, or hope that it will actually be viable off the blocks.
The hardest decisions are from M you always have to be careful for deferred maintenance and disappearing staff.
Sorry, but the intelligence built into the protective relays is pretty robust. The problem is that we isolate elements to protect that element. This strategy is predicated on a single action not having a cascading effect-- namely having sufficient spinning reserve that generation capacity is not overloaded. The same holds true for major transmission lines, but increased spinning reserve helps to add generation capacity closer to the load.
The alternative, which we do for campus-scale systems is to have means to quickly isolate blocks of load well downstream of the generation and transmission topology. We build failure-on-failure, so we know that if communications links are severed 25% of the load will drop.
The problem is that there are finite limits to how well this scales; latency is critical. You have to ensure that all communications are viable all the time. This is the challenge I have with the "smart grid" talk; deep load pickup/shed capabilities are interesting from a scheduling perspective (flatting demand profile), but do very little in emergency response.
Ten years ago when I last toured an ISO's command center, they were able to project load to within 0.5% 24 hours in advance. Granted, spinning reserve was higher back then, but the fundamental logic hasn't changed much.
So I am lost as to what the smart grid is actually supposed to do, aside from a fancy version of automated demand-response. It wouldn't be fast enough to actually function as "protection".
Transmission protection relays will usually have a fiber optic link to establish distance to fault, direction of fault, and a number of other issues. This protects the transmission line.
The generation station might have two or three transmission lines coming out of it. If it needs to shed load, it will dump one of those lines. But, the ISO (independent system operator?) coordinates all the generators, so you are likely to have the same load just moved to the other two lines. Ultimately, this forces the generator to shut down completely.
What you could theoretically do is automate the role of the ISO to coordinate generation and demand in real time. If you need to shed 5% of the load, it is very hard to do it right now in real time. That is the failure that most of these attacks try to hit, as it is easy to cause a cascading failure if transmission is limited.
Remote access and e-mail notifications more often drive the internet connections we have seen. When facility engineering is out-sourced, it becomes even more complicated, because there is fundamental conflict in the way the contracts are written-- the Owner might require all security go through them, but they don't allow the facility engineers to be on their network.
Usually you end up with a DSL connection and a "firewall router." Usually it is just a monitoring network, but control seems to creep in more each month.
It is easy to say that a data center needs its own source of standby power, but there are a lot of industries where a one minute outage causes a six to twelve-hour restart time, and the margins are far too thin to support the additional generation infrastructure.
I know a few people putting in multi-megawatt fuel cell systems, but these have the same restart-time problem that the utility grid has, which is ultimately the problem.
The easiest fix at a utility scale is to increase the amount of spinning reserve so that causing a cascading failure requires control of multiple generating facilities. Once you trip a facility offline, re-start times are just way too long, especially given emmissions controls.
I have trouble believing that the "smart grid" really solves this, although you can do some things with networked protection strategies and more selective load shed.
Dryers and ovens are run at either 208 or 240V in the US; we use a center neutral, so residential services still have 240V power available for large appliances.
Moreover, the mechanical strength of the conductors limits practical sizes to #12-14, eliminating most benefit of higher voltage. With 230V, you need to be careful to have enough fault current to trip a breaker.
Also, the US uses 277/480V power for anything that needs real oomph, which both limits the amount of transformations, and provides bigger bulk power and higher torque, all while keeping common plugs at a safe, non-arcing voltage.
That is actually really cool. It does a lot to make up for all the problems in the UK design... I wonder if it would pass an electrical safety testing.
A license extends rights that are not otherwise available through copyright law. The logic is that you don't have the right to copy the disc contents onto your computer without said license. From a legal perspective, the shrink-wrap license is here to stay (sadly).
Interestingly, a license cannot restrict rights of the purchaser that would otherwise be available under copyright law; it can only grant new rights.
What is wrong is that you can't return opened software when you do not agree with the terms. You are not prevented from selling the software on under first-sale rights though.
I second the comments on Buffalo. They die a day out of warranty. Very difficult to do snapshot backups. Grossly under powered. No hope of actually being able to rebuild an array.
For the money, it is hard to find anything adequately reliable without a diy job.
What right do you have to go to grad school straight after undergrad? Do you think that impacts your risk profile vs someone who works for a few years in between to pay down loans and decide that they really want to be doing what their masters is in?
Far too many people in your situation go to grad school for the safety net. Doctors are an obvious exception, but their pay offsets it reasonably well.
Does your education have a positive value proposition for you? What multiplier of expected first year salary did you have to take out?
While my education was a great value ($500/semester with an in-state reciprocity rate when I started 20 years ago), the student loan was one of two things that kept me in my field long enough to get my PE, and to eventually realize that I love the field. I have heard similar stories from others.
While it is a burden, it is a real first lesson of responsibility in your career. Also, with the interest being deductable it is me who should be complaining for subsidizing you if anything.
What is the maximum garnishment? The rate is high because defaults are high because people take out too much in loans because tuition is high and it is easy to do.
If you think 8% is high then the real world is going to be even more scary: jmbo mortgages are really 6%+, small business loans about 10% if you can get them, and loans for purchasing furniture 18%. Oh, and then there are credit cards.
Because this does not lead to research into more efficient technologies. When Title-24 energy efficiency requirements first came to be, they allowed 1.5 watts/square foot of lighting for office space. Over the next 20 years, that was ratcheted down to 1.0 w/sf today. You couldn't have gotten that low 20 years ago, but now it is possible through better lamps and fixtures.
Today, when I go into a store where all the lighting is incandescent track lights (obviously not in California), I am amazed that the owner wouldn't take it on themselves to improve the lighting and dramatically reduce their energy costs and improve merchandising quality, all with a payback of less than a year.
People don't put two and two together on these things without a push.
Dice looks just as bad as Monster for my field (Architectural Engineering); one query has 14 hits in the Los Angeles market, of which two were direct, and the remainder were squatted by recruiters.
One of the direct positions is widely advertised, leaving little added value for Dice.
I never realized how hard it is to cull applicants until owning my own business.
There are a few different components at play: profit, initial cash, payment cycles, and risk profile. Billing rate (for us at least) is about 3x nominal salary, and the employer's gross profit margin is roughly equal to nominal salary.
If you have cash to start two employees today, you hope to generate cash flow with them in 3 months. You can use that same "at risk" capital to hire two more people after that 3 months. If you get a 20% cut in pay, you can obviously afford (almost) half of a third person initially. By taking risk on that third person based on half their salary, you can make it work.
You expect cash to come in after month 3 on the first three people. By accepting risk on the A/R, you could hire three more people a month early. (The higher profit margin means that you could afford 4 people in month three alternately, but you lose 3 man-months of profit, so the risk should pay off.) In month 5, you have enough cash to hire three additional people, and in month 8-9 you have fully recovered your at-risk capital, and after three more months you can pay them out for their contribution.
Of course, you have to have enough business to make it work, access to enough at risk capital, and clients that pay invoices in a reasonable timeframe. You also need to be able to hire people that can hit the ground running.
Slashdot Mirror
Linky. The roof came off over the Pacific.
The 787 program was started 2-3 years after the A380, and a significantly more difficult engineering task than the 380. Airbus had the same kinds of problems Boeing did in their supply chain. Airbus had to cancel their freighter version. Both companies have massive government subsidies that the other side's government proclaim are illegal.
Boeing was obsessed with the Sonic Cruiser before the 787, which is why they ended up behind the game compared to Airbus.
Personally, my favorite plane to fly will always be the upper deck of a 747, as it creates an intimate space where you forget about all the other people on the plane. The 380 is more like to 777's stacked on top of each other. It might make sense in some markets but it is hard to believe that it will be the long-term solution.
Time will tell...
Ground temperature at the surface to a reasonable depth (not sure exactly what that is) is equal to the average annual temperature. The problem is doing horizontal pipes below the frost line vs. vertical bores spaced adequately apart.
We looked at doing a geothermal project for a limestone mine to be converted to a data center, but it wasn't practical as they filled it in to reduce flooding risk and chamber height. Best approach is heating/cooling the aquifer, but that has similar problems at a macro-scale.
The real issue is if you want to be in a technical role for life, or if you would like to transition to management. The initial change from technical to managerial positions means a sacrifice, but long-term there are significant benefits.
In my field, people below the line get paid overtime, and above the line are pure salary. Many people get 10-15% overtime, and only a 5-7% pay increase for crossing the line, thus taking a pay cut in the process. Within a year or two they usually make up the gap and then some, as they are eligible for higher bonuses instead.
If you think you are just going to become a worthless PHB and that has no interest for you, by all means pass though...
Not enough error correction to accommodate one good scratch. Detecting an error and recovering from it can be very different things.
If you want to see computers saving hospitals money, get a physical at Bumrungrad Hospital in Bangkok. The system is multi-lingual, and you get results almost immediately. I don't know that a legacy hospital can pull off the same things, but it is truly amazing to see.
Compared to my physician's scribbles on pre-printed form letters, it really does a lot to make it look like the doctor knows what they are doing and communicate effectively.
The poster blamed some nebulous reasoning; we don't know what Apple really said.
Good enough for missiles, not good enough for Chicago (or Honolulu)...
Even if you are going co-lo, you are best off hiring a pro to help you. (Disclaimer: I am a pro.) The checklist approach is a good starting point, but far too often you will miss site-specific issues.
How old are the UPS batteries? Where are the preventative maintenance logs? What happens if the facility is expanded? Are there any single points of failure you need to understand? Is it 6N/5 or N+1? Are there lease terms that you need to be careful with?
We have done big deals with large customers, and despite having a good internal team, good legal experts, and it is always amazing what little things come up that are really critical to addressing properly to protect liability.
Purchasing a facility is often easier; you either know the facility is a write-off, has some level of residual value, or hope that it will actually be viable off the blocks.
The hardest decisions are from M you always have to be careful for deferred maintenance and disappearing staff.
Sorry, but the intelligence built into the protective relays is pretty robust. The problem is that we isolate elements to protect that element. This strategy is predicated on a single action not having a cascading effect-- namely having sufficient spinning reserve that generation capacity is not overloaded. The same holds true for major transmission lines, but increased spinning reserve helps to add generation capacity closer to the load.
The alternative, which we do for campus-scale systems is to have means to quickly isolate blocks of load well downstream of the generation and transmission topology. We build failure-on-failure, so we know that if communications links are severed 25% of the load will drop.
The problem is that there are finite limits to how well this scales; latency is critical. You have to ensure that all communications are viable all the time. This is the challenge I have with the "smart grid" talk; deep load pickup/shed capabilities are interesting from a scheduling perspective (flatting demand profile), but do very little in emergency response.
Ten years ago when I last toured an ISO's command center, they were able to project load to within 0.5% 24 hours in advance. Granted, spinning reserve was higher back then, but the fundamental logic hasn't changed much.
So I am lost as to what the smart grid is actually supposed to do, aside from a fancy version of automated demand-response. It wouldn't be fast enough to actually function as "protection".
Transmission protection relays will usually have a fiber optic link to establish distance to fault, direction of fault, and a number of other issues. This protects the transmission line.
The generation station might have two or three transmission lines coming out of it. If it needs to shed load, it will dump one of those lines. But, the ISO (independent system operator?) coordinates all the generators, so you are likely to have the same load just moved to the other two lines. Ultimately, this forces the generator to shut down completely.
What you could theoretically do is automate the role of the ISO to coordinate generation and demand in real time. If you need to shed 5% of the load, it is very hard to do it right now in real time. That is the failure that most of these attacks try to hit, as it is easy to cause a cascading failure if transmission is limited.
Remote access and e-mail notifications more often drive the internet connections we have seen. When facility engineering is out-sourced, it becomes even more complicated, because there is fundamental conflict in the way the contracts are written-- the Owner might require all security go through them, but they don't allow the facility engineers to be on their network.
Usually you end up with a DSL connection and a "firewall router." Usually it is just a monitoring network, but control seems to creep in more each month.
It is easy to say that a data center needs its own source of standby power, but there are a lot of industries where a one minute outage causes a six to twelve-hour restart time, and the margins are far too thin to support the additional generation infrastructure.
I know a few people putting in multi-megawatt fuel cell systems, but these have the same restart-time problem that the utility grid has, which is ultimately the problem.
The easiest fix at a utility scale is to increase the amount of spinning reserve so that causing a cascading failure requires control of multiple generating facilities. Once you trip a facility offline, re-start times are just way too long, especially given emmissions controls.
I have trouble believing that the "smart grid" really solves this, although you can do some things with networked protection strategies and more selective load shed.
Dryers and ovens are run at either 208 or 240V in the US; we use a center neutral, so residential services still have 240V power available for large appliances.
Moreover, the mechanical strength of the conductors limits practical sizes to #12-14, eliminating most benefit of higher voltage. With 230V, you need to be careful to have enough fault current to trip a breaker.
Also, the US uses 277/480V power for anything that needs real oomph, which both limits the amount of transformations, and provides bigger bulk power and higher torque, all while keeping common plugs at a safe, non-arcing voltage.
That is actually really cool. It does a lot to make up for all the problems in the UK design... I wonder if it would pass an electrical safety testing.
A license extends rights that are not otherwise available through copyright law. The logic is that you don't have the right to copy the disc contents onto your computer without said license. From a legal perspective, the shrink-wrap license is here to stay (sadly).
Interestingly, a license cannot restrict rights of the purchaser that would otherwise be available under copyright law; it can only grant new rights.
What is wrong is that you can't return opened software when you do not agree with the terms. You are not prevented from selling the software on under first-sale rights though.
I second the comments on Buffalo. They die a day out of warranty. Very difficult to do snapshot backups. Grossly under powered. No hope of actually being able to rebuild an array.
For the money, it is hard to find anything adequately reliable without a diy job.
What right do you have to go to grad school straight after undergrad? Do you think that impacts your risk profile vs someone who works for a few years in between to pay down loans and decide that they really want to be doing what their masters is in?
Far too many people in your situation go to grad school for the safety net. Doctors are an obvious exception, but their pay offsets it reasonably well.
Does your education have a positive value proposition for you? What multiplier of expected first year salary did you have to take out?
While my education was a great value ($500/semester with an in-state reciprocity rate when I started 20 years ago), the student loan was one of two things that kept me in my field long enough to get my PE, and to eventually realize that I love the field. I have heard similar stories from others.
While it is a burden, it is a real first lesson of responsibility in your career. Also, with the interest being deductable it is me who should be complaining for subsidizing you if anything.
What is the maximum garnishment? The rate is high because defaults are high because people take out too much in loans because tuition is high and it is easy to do.
If you think 8% is high then the real world is going to be even more scary: jmbo mortgages are really 6%+, small business loans about 10% if you can get them, and loans for purchasing furniture 18%. Oh, and then there are credit cards.
Because this does not lead to research into more efficient technologies. When Title-24 energy efficiency requirements first came to be, they allowed 1.5 watts/square foot of lighting for office space. Over the next 20 years, that was ratcheted down to 1.0 w/sf today. You couldn't have gotten that low 20 years ago, but now it is possible through better lamps and fixtures.
Today, when I go into a store where all the lighting is incandescent track lights (obviously not in California), I am amazed that the owner wouldn't take it on themselves to improve the lighting and dramatically reduce their energy costs and improve merchandising quality, all with a payback of less than a year.
People don't put two and two together on these things without a push.
Dice looks just as bad as Monster for my field (Architectural Engineering); one query has 14 hits in the Los Angeles market, of which two were direct, and the remainder were squatted by recruiters.
One of the direct positions is widely advertised, leaving little added value for Dice.
I never realized how hard it is to cull applicants until owning my own business.
Too hard to train toddlers.
There are a few different components at play: profit, initial cash, payment cycles, and risk profile. Billing rate (for us at least) is about 3x nominal salary, and the employer's gross profit margin is roughly equal to nominal salary.
If you have cash to start two employees today, you hope to generate cash flow with them in 3 months. You can use that same "at risk" capital to hire two more people after that 3 months. If you get a 20% cut in pay, you can obviously afford (almost) half of a third person initially. By taking risk on that third person based on half their salary, you can make it work.
You expect cash to come in after month 3 on the first three people. By accepting risk on the A/R, you could hire three more people a month early. (The higher profit margin means that you could afford 4 people in month three alternately, but you lose 3 man-months of profit, so the risk should pay off.) In month 5, you have enough cash to hire three additional people, and in month 8-9 you have fully recovered your at-risk capital, and after three more months you can pay them out for their contribution.
Of course, you have to have enough business to make it work, access to enough at risk capital, and clients that pay invoices in a reasonable timeframe. You also need to be able to hire people that can hit the ground running.