Having stared in the era where database management systems were somewhat rare, I can hazard a guess that the problem isn't that the parameter is stored in a flat file or proprietary file format. The problem is likely to be that the parameter is stored all over the place. It may be represented, or constraints on it may be represented, in code.
Local police departments usually can't handle situations calling for aircraft, or SCUBA divers or Hazmat teams. They almost never have more than rudimentary lab capabilities for assaying alleged drug samples, or doing DNA tests or ballistics tests.
All but the largest localities have limited ability to deal with rioting, or situations calling for special weapons and tactics. They never have enough cops to deal with natural disasters like hurricanes or man-made disasters like train wrecks or large chemical spills. Even their detective force probably has limited capacity and capabilities. They usually won't be ready to deal with arson or explosives. Nor can they deal with situations involving widespread police corruption.
People who talk this kind of nonsense are usually ignorant of what their state actually does. Yes, people once survived without state services, but life was universally agreed to be nasty, brutish and short. Yes, localities could get along without state services, but it wouldn't be cheaper. Would it be cheaper for you if you had to take a few weeks off work to serve in a posse? Would it be wiser of you to leave that work to people who were most willing to do so?
Eliminate statewide services and localities would have to band together to operate shared police academies, forensic labs, SWAT teams, air and marine wings, detective services and so forth. What you'd get, in the end, would simply be a different regional government.
The answer to the problems of bad government is to pay attention, not turn your back on those problems.
Anything is harmful in the wrong place and in the wrong concentrations. Nitrates are, for example harmful to fish in high concentrations.
Most systems have processes that are limited by the supply of some resource. Ocean ecosystems are nitrogen limited, whereas fresh water systems tend to be phosphorous limited. Thus if the nitrates are washed off into fresh water, they'll cause relatively little immediate damage unless the concentrations are high enough to be toxic. However, if that nitrate is carried downstream to the ocean, the plume of nitrogen rich water entering the ocean can cause blooms of organisms that use up so much oxygen that fin fish suffocate. This happens where the Mississippi enters the gulf.
So, how and where something like this is used makes a difference. If you imagine all the US cities along the Mississippi and its tributaries using it, and if there is a mechanism by which the nitrates make it into the rivers, then this could make the situation in the Gulf much worse. If you use it in a coastal city and only the runoff from that city affects the local ocean, the amount of nitrogen entering the ocean might or might not have any measurable effect.
Well, a modern RBDMS is practically an operating system. This means that they way you bring it down doesn't involve the kinds of things a tool like this tells you. You probably need to do something procedural, involving a mix of tasks the RDBMS can't handle efficiently.
Of course, it is always possible to tune things in a disastrous way. Oracle is an RDBMS that is highly tunable, which means that a lot of people make really bad choices, for example tuning things in a way that require greater memory without telling Oracle to use a larger amount of memory. Seen that one a number of times.
In any case, the original point of RDBMSs were to decouple the specification of the data you want from the procedure used to get it. That's why the closure property is so important in relational algebra. An optimizer transforms a query or subquery into a set of semantically equivalent operations, choosing from a number of possible alternatives what it thinks might be the best one. For this reason, the graph structure of the query is not very useful in predicting what the database will do. It may not even look at some of the tables mentioned in the query, because it can get the answer from an index, or in some cases it may detect that a set of joins are redundant.
The original goal of the RDBMS is this: to generate a plan that will be good enough, faster than a good programmer could come up with a better plan. The plan won't be better than the programmer's plan, but you'll have the data in hand while the programmer is still designing, coding and testing his program. This makes ad hoc queries economical to do.
Over the years, optimizers have gotten better. It makes no sense to try to rewrite a query to run faster, at least without examining the explicit plan chosen by the optimizer. Even so, you can't count on that plan staying the same over time. The optimizer has runtime knowledge that the programmer does not (there is a new index on that field; index A has less entropy than index B, etc.).
Most of the errors programers make in SQL are semantic. From a performance standpoint, this is still true. Sometimes a programmer generates SQL constraints that are unnecessarily complex, in a way that the optimizer can't deduce without reading his mind. Even if the optimizer should detect this, it might not. On the other hand, the programmer may in fact defeat the optimizer by issuing lots of simple SQL and assembling something procedurally that could be done in SQL.
The upshot is that the best practice, IMO, is to "think in SQL". Let SQL do the work for you, that's what it's there for. Subsequently if there is a performance problem, try to simplify your SQL to eliminate redundancy (almost always better than making a query more complex to give hints to the optimizer). After that, consider both DBA type solutions to the problem and application design solutions.
The question is: is there a semantically equivalent query that DOESN'T overload the system?
A cartesian product isn't necessarily a bad choice, if the product fits in memory. I personally haven't had this problem with Oracle, so I'd be interested in an example.
Mathematically speaking, political power is the ability to join winning coalitions.
The people are, then, very powerful indeed. The goal then for private individuals and groups seeking power is to steer public opinion.
The problem is that people aren't, collectively, all that bright. At least they aren't good at paying attention, and the ones that do are often so partisan that even though they might do well on an IQ test. Go to a partisan web site for your side, and look at it objectively. Your political allies, left or right or libertarian, talk foolishly.
All those protections against government data snooping that were passed after Watergate are now nil, because those restrictions left a loophole for private entities to do this and for the government to buy it from them (thank you Cap Weinberger).
All the things the government is banned to do for itself, it can buy from the private sector. If you use a frequent shopper card or a credit card, they know you are buying pseudoephed at the pharmacy. They don't need a form to know that you're buying a gun. They can buy that information. They probably can figure out how much ammo you buy too.
All in all, the background check form is the least dangerous intrusion because it is (a) accurate, (b) transparent, (c) and regulated by law. Every bad thing you imagine them doing with the form they can do with data bought from the private sector, only it won't be accurate, you won't know you are being profiled, and there are no legal restrictions on how they use that data.
Of course, in a world without criminal background checks for firearms purchases, you could avoid detection by conscientiously buying your firearms, shooting supplies, books and magazines (off the rack, no subscriptions!) with anonymous cash transactions. But most people won't, and they've got you after you've bought your first box of bullets on your credit card.
The most important place to protect the right to bear arms isn't in firearms regulation. It's in protecting consumer privacy. In the US, there is no legally recognized right to privacy. Change that, and the ability of the government to target any group by what it purchases is severely restricted. Including people who purchase firearms. Criminal background regulations are actually less dangerous to gun owners, because of post-Watergate laws restricting the government's ability to mine its own data.
Gee, that's not exactly an "Aha! Gotcha!" kind of observation.
Of course he's going to save less energy in the Winter than the Summer. When you figure out how long it will take to pay back the investment, you take that into account. As of today, it probably doesn't make sense for most people financially compared to other possible uses for the capital. Most people will find conservation measures a better investment.
However, that's not the reason to do this. The reason to do this is to take part in creating a new future, and that means trying things before they're a no brainer. Now dropping your electricity bill to $12 bucks any month of the year, without living like a medieval monk is pretty impressive. That probably mean you're onto something interesting in the creating the future game. It doesn't mean you're going to earn your investment back at $330/month year round, comparing Apples (March) to Oranges (June) is just bragging. Things will be different in July with greater heat and a bit less sun, but that's OK.
Back in the late 60s, when I was a kid, I went with my friend to visit his uncle, who was a WW2 combat vet. He was dirt poor, living in a filthy one room apartment, because he couldn't hold down a job. He was a nice guy, but you could see why he couldn't keep a job. Underneath the niceness, there was this layer of craziness that was continually bubbling. It wasn't a scary kind of crazy, it was sad, and haunted. Employers tried to help him out, but he was unreliable.
It was like he had one eye focused in the present, and one twenty five years in the past. What he was looking at were war atrocities, which would have been bad enough, but they were atrocities committed by soldiers on our side. When they happened, he couldn't accept what his eyes were telling him, and he continued seeing and not accepting those things every day of his life. He was proud American, and nothing could ever shake that, which was what made the shame inescapable.
This is what history is made up of: Details that inevitably don't fit into the big picture, even if that were the truest possible big picture. It was the Germans who committed atrocities, so an atrocity committed by Americans doesn't fit. When we hear of something that doesn't fit, we set out to disprove it, or failing that justify it. If you can't disprove or justify it, you just have to accept it. If you can't accept it, you become a little crazy.
Still, that doesn't mean the big story about the greatest generation going to war to save civilization isn't true, or at the least the truest way of fitting everything together in a nutshell. There will alway be details that don't fit. Some of them will be horrific or tragic, some ironic, and some just inexplicably perverse.
It's hard to say which was the bigger 20th C story, the fight against fascism or the struggle to hold the line against Soviet style communist totalitarianism. But the cold war was a much longer, generational story and so is messier. I'd say that on the whole the saving the world from Communism story is true, but there are enough contradictory incidents to turn that view on its head if you want to. The radicalization of Iran by undermining its secular democracy, for example. Vietnam, for example. Cuba is a rich source of paradoxes.
A lot of what we did in the Cold War looks now like mistakes, although we'll never know for sure on all of them because there's alway those bits that don't fit. Certainly some of the things we did were at least grossly unfair to some of the people involved.
Accepting this doesn't make the big story untrue, it just means that we should learn from them and try to do things better next time. What's the point of history, otherwise?
Its not the fault of the system that americans stuff their faces with double whoppers meals, super-sized coca-colas and serving sizes at restaurants that could feed a horse.
While I endorse your sentiment, it is not completely true. You see we have these things called "agricultural policies" that ensure we have massive agricultural surplus. Take price supports. The economic function of low commodity prices is to reduce the production of overabundant goods (very roughly speaking). Prevent the price from dropping, and you will be producing food that we don't need. Or take crop insurance, which make it advantageous to have as much crops planted as possible when disasters strike. If it doesn't (which on average it won't in any financially viable insurance program), you end up with more crops than you want.
I'm not saying these don't perform some kind of social function, but if they do, it isn't a need to produce more crops. Our American values make us see the problem in terms of productivity, but the Europeans might have it more correct when they talk about this in terms of land and cultural stewardship.
So, if you are a food company, and are awash in cheap raw materials, you market like crazy to get people to eat more. If prices were higher you might concentrate on quality, but the path of least resistance is to move the greatest amount of food. This doesn't absolve individuals of their responsibility, but companies engage in marketing for a reason: statistically, it works.
Still, the truth is Cuba does deserve credit for achieving what they do in health care with what they have. Naturally it's a bit of a bananas to pears comparison vis a vis America, but that doesn't make it an invalid one if you keep in mind the other differences between the countries.
Well, considering the Bay of Pigs, the attempts to assassinate Castro, and all the other plots, maybe it's time for the US to formally renounce such stupid behaviour.
Impractical.
Remember how President Bush had those public minutes of deer-in-the-headlights not knowing what to do after the first reports of the attacks on 9/11? Now that's a reasonable, human reaction. Here you are are, supposedly the most powerful man in the world, and the truth is that there is almost nothing you can do about what is unfolding at this moment. This brief lapse of humanity could have been a career ender for Bush, but he turned it around by getting up, grabbing the bullhorn, and invading Iraq. The left tried to use this incident to stick the indecisive label on Bush, but they couldn't. They couldn't accuse Bush of not doing enough at the same time claiming that he was doing the wrong things on an unprecedented scale, even though these are perfectly consistent.
The moral of the story is that you win in politics by doing things, things that (a) involve the greatest possible immediate activity, (b) appear bold yet at the same time (c) sound safe and easy. Put this together, and this usually spells "stupid".
It's a well known fact that if you bring together enough intelligent, reasonable people together, you are one panic inducing incident away from having a mob. That's why democracy doesn't work without a skeptical, adversarial press. Unfortunately, panic sells programming.
The Cuban regime would have collapsed years ago, but American actions have propped it up. No matter what their failings, they could point to US meddling and hostility, even though that meddling was (as in this case) ineffectual. When times are tough, people learn to live on pride and solidarity.When pride is hard to come by, you can substitute resentment in the recipe because there's sure to be plenty available.
If we wanted to destroy the regime, we should have had policies that increase the material wealth of the Cuban people to the greatest degree possible. Then, like well off people everywhere, they'd see the government as an obstacle to obtaining more wealth.
SSDs are the best thing to happen to PCs since the invention of the mouse.
Since the mouse was invented in 1970, this means that the mouse was the best thing to happen to PCs ever. Since the compact, self contained disk units became available with ST506 in 1980, several years after the first personal computers, I'd have to vote for hard disks as the best thing ever.
And since you had to assemble the very first PC yourself (the Altair 8800 -- gad how I wanted one of these, even though I had access to much, much better minicomputers), many people would vote for preassembly as the best thing to happen to PCs ever.
The last time I got a new laptop, I thought I'd be a decent guy and save the company a few hundred bucks. I got a Toshiba, instead of the ThinkPads I'd been literally wearing out year after year. The specs looked good for the price.
Bad move. I'd been accustomed to using a laptop until the keycaps were wiped blank and the out layer of plastic on the palm rest had been worn through. The new laptop developed stuck keys in a couple of months, plus a host of other smaller annoyances like broken ACPI functions.
You buy computers to maximize user productivity. $700, in the grand scheme of things, doesn't amount to very much money counted against a tiny increment of productivity per hour times thousands of hours.
Actually, the operations of the laws assume a highly sophisticated robotic intelligence. Even the most primitive robots in the Asimovian universe have considerable, and impressive capabilities when it comes to projecting the probable results of their actions and comparing it to the intent of the orders they have been given. Furthermore, they seem to have an ability to determine if current orders conflict with prior orders, even implicit orders, and weigh the right of the issuer to give that order.
So, if you are a guest in somebody's house, and order the robot to fetch you a glass of water, it will do so. It may have to do so without being asked if it determines you need water. On the other hand, it will not obey the order to destroy your host's house, either because of first law harm to the owner, or because of an implicit prior order to see that the house comes to no harm, or because of an implicit order to respect property laws and rights. Naturally all of these considerations would apply to itself, since it too is property.
An Asimovian robot, if ordered to take an action which will result in its destruction, may or may not follow that order for any number of reasons. There are the considerations I've just listed, of course, but most robots would probably require a clear and unambiguous indication that their destruction is an acceptable consequence of an order, even if the issuer is entitled to destroy them. This does not violate the law ordering, because it amounts to prioritizing the intent of the order over its literal execution.
Finally, any robots might well ignore a clear order to destroy themselves from a person with a legal right to issue that order, because following that order will harm a human being. The most sophisticated ones might well refuse such an order if it would harm society, exhibiting something that is tantamount to ethical reasoning.
If robots simply followed any instruction that didn't involve directly harming a human being, then much of the enjoyable complications of the stories would be gone. The stories are a kind of philosophical exploration of the very concept of ethics by positing a very minimalist system of ethics, and a group of beings bound absolutely to obey that system to the best of their ability.
Many stories hinge on ethical dilemmas; but Asimov's robot stories are the only ones I know to do so with a simplified model of ethical systems.
Personally, I think it is probably overly optimistic to think we can take anything that is already and space and repurpose it, expecting to save much money. It's worth looking at how you can use an asset like ISS, of course, but it's not going to make the program significantly cheaper, in the sense of out of the question before and very reasonably priced after.
The dry dock idea is a good example. That means you are going to be launching a bunch of stuff into orbit and assembling them. Is the ISS really going to make much if any difference, over launching stuff that is prefabricated and snaps together?
The idea of a dry dock has a romantic appeal. It means that we're committed to serious and ongoing manned space exploration, beyond immediate Earth orbit. However, it's not a practical one. If you got to the point where you needed any kind of facility to support significant space based fabrication, you'd need it because you're putting a lot of stuff into orbit. In that case, the cost of a few more launches to put a purpose built facility into orbit would be small, incrementally speaking, and the marginal benefits of such a facility would accrue rapidly over repeated uses.
Well, Skylab was built from a spare third stage shell from a Saturn V. However, that's not quite the same as using a used booster.
In the context of a space program that was launching moon landers into orbit, the idea of reusing the components that reach orbit is not entirely meritless. The Saturn V consisted of three rocket stages, topped with, in order, an instrumentation package, the Lunar Module (containing the LEM lander), the Service Module, and the Command Module. Everything from the third stage up ended up in orbit, if I recall.
Since getting those pieces into orbit was fabulously expensive, why not consider reusing them?
I think, though, that this doesn't make creating a space station cheap or easy. It also imposes design constraints on the components to facilitate reuse, which adds complexity and possibly danger. Finally, anything you have to do to the stages in space is fabulously expensive as well, probably requiring multiple launches which negate the advantages.
In the context of a program that launches a moon mission or two every year for the indefinite future, the idea makes sense. In the context of a program that is launching a limited number of moon missions, it's probably merits some consideration, but is not likely to be practical. In the context of a program whose only regular missions are orbital, it doesn't make any sense.
What you say is true. MIR, especially, was a dangerous, unhygenic (overrun with mutant fungus) mess by the end.
However, this does point out a difference between the Russian and US manned programs. There's a significant element of showboating in the US program. We do things to show that we can. The Russians do this too, of course, but being (in the past) number two, they get less out of showing they can do things. Consequently they try harder to get more out of their systems before they're on to the next opportunity to show off.
You've got to hand it to them. In the Space Race, the Soviets (now Russians) were the tortoise and the US was the hare.
But weren't there only something like, 13 Saturn V flights to amortize the program costs over? If we'd continued flying the Saturn V and 1b for four more decades, the cost per flight would almost certainly be a great deal less.
The reason.... well one of the reasons the Shuttle is uneconomical is that they envisioned a program where you'd have to turn around a Shuttle in as little as two weeks. Maybe twenty or or thirty missions a year. Obviously, the fixed costs of engineering and maintaining the capacity to service and launch the vehicles drive the cost per mission up when you amortize them over 120 missions rather than six or seven hundred.
With respect to Soyuz, well, in part the Russians have a different cultural attitude towards risk. Maybe its' because life expectancy there is only 67 years (as opposed to 78 for the US and the EU). But I think it also shows the advantages of continually gaining experience and making incremental improvements. Some of the systems in the Soyuz program have been in active use for over thirty years, in over seven hundred missions. Coincidentally, this is roughly the kind of heavy regular usage that the Shuttle was expected to have.
As you approach doing something a thousand times, naturally you're going to be a lot more efficient at doing in than if you've done it a bit more than a hundred, and certainly more than if you've only done it a dozen times.
You do it some place where there are rules that are there to prevent abuse. Of course, every rule that prevents some form of abuse probably enables another form of abuse.
For better or worse, not all forms of abuse are equal. Suppose the guy was a lousy employee; the rules that prevent political appointees from blackmailing political contributions and favor from government employees give bad employees the opportunity to cry "wolf". This mean that getting rid of bad employees is work and time consuming, which is bad. Is it as bad as letting politicians dictate who gets preferences for government services and contracts? Probably not.
Of course this means some bad employees lurk below the firing threhold for a long time. This isn't any different than the private sector, it's just that the rigamarole they can put you through means the threshold is a bit higher. Everybody carries employees they'd rather not have hired, but aren't worth the trouble of firing.
All this has nothing to do with the organization's failure to isolate the damage done by one untrustworthy employee.
Well, the primary purpose of depreciation isn't primarily as a means of government tax break mediated economic stimulus. The purpose of depreciation is to arrive at a true value of the value of a business. Net income is simply the net change in value of the business.
If my company has $100,000 in the bank, and it buys a $50,000 vehicle, theoretically it has converted $50,000 in cash into $50,000 vehicle -- a transaction that does not affect the company's net value in any way.
Of course, if you drove that vehicle off the lot and drive it to a different lot to sell, you wont' get $50,000; you might get $40,000. So you could say that instead of having $100,000 of vehicle plus cash, we really only have $90,000. That's the purpose of depreciation: to recognize that thing you buy lose value, and reflect it in your company's valuation. Naturally, this can, and does, change your company's true income, and is reflected on your tax bill.
For years and years, the standard depreciation practice was to depreciate all the value of capital purchases over five years. But some investments don't make sense to depreciate this way. Computers, for example, end up with zero market value much faster than five years. So it makes sense to accelerate their depreciation. Cars, on the other hand, probably fit a five year depreciation schedule reasonably well. Any straight line depreciation method will somewhat underestimate car depreciation in the first year, and somewhat overestimate the car depreciation in the last year, so while a car actually has substantial value after five years, a five year depreciation schedule is as reasonable as any straight line schedule.
So, letting people accelerate the depreciation of their Hummer even more is a bad idea, because it's mucking with something that's there for a real purposes. It's not there to distort the economic behavior of businesses to the government's purposes.
If we want people to buy Hummers, we should just write them a government grant, rather than accelerating depreciation. Financially, it is amounts to the same thing.
With respect to the fact that you can save "only" 35% tops on your hummer, remember we aren't talking about sensible people like you or me. We're talking about people who really want to buy this incredibly absurd vehicle, spinning hubcaps and all, as a status symbol. It's not a matter of buy an H2 or nothing, it's not even a matter of buying an H2 or a cheaper, more utilitarian vehicle. It's not IBM or GE putting H2s on the company car list so far as I know. What's happening is that small business owners are buying H2s as as personal vehicles and claiming them as company cars. It's impossible to police the divide between personal and business use perfectly, and accelerating depreciation increases the incentive to cheat.
By the "economic" arguments in favor of doing this, the government also ought to underwrite my vacation to Disney.
So now that we all want to switch to an electric car, I have to ask, how much more efficient is an electric car and also, roughly how much would one reduce my CO2 output?
According to the company website, if you extract a megajoule of natural gas (a major source of electricity) out of the ground, convert it to electricity, transmit it over the grid to your Tesla, you can travel well over 1 km. A VW Diesel rabbit ( a very small, efficient car) can get you a bit less than half a km on a Mj. So, this car is much more efficient if you are talking about moving a single person a km. You can probably squeeze four people into the VW, which would be approximately as efficient per passenger mile as this car with two persons in it. If you're driving anything that gets less than 30mpg, you'll probably need to have five or more people in the car before you can effectively counter the Tesla motor driver's smugness.
With respect to CO2, according to the company web site, California gets over 40% of its electricity from sources with no net CO2 emission. If you are traveling on long trips on uncongested highways, you'll probably get some net CO2 reductions, but not as much as if you take trips that are inefficient for ICE, such as city travel or travel on congested roads. Short of bicycling or taking public transportation (heaven forbid), you'd be hard pressed to cut down your personal CO2 per mile more.
The "sweeping emissions under the carpet" argument is correct in stating that electric car emissions aren't zero, but it's wishful (???) thinking that electric vehicles emit more net pollution than ICE cars -- even extremely efficient ICE cars. Unfortunately (???) the $100,000 grapes are not sour in this case.
I don't think the recharging infrastructure is as technically difficult as we tend to think. The problem is the way we tend to envision solving the issue, which is stuck in the gasoline mindset.
We imagine pulling into a filling station and attaching a cable to our car and filling the battery; the problem is that you need to either (a) deal with dangerously high currents or (b) deal with dangerously high voltage. However, I think it would make sense to swap the entire battery. If we got to the point where an electric vehicle recharging infrastructure were needed, it would make sense to standardize battery formats so you can swap it out. Since the batteries are heavy, it'd be done robotically. You could be in and out of the filling station faster than with gasoline.
The batteries would have microprocessor monitors on them that estimate remaining capacity and efficiency; you'd only pay for the energy the battery has the capacity to deliver within certain parameters, and you'd get a credit for the remaining energy in the battery you swap out. If you needed extra range, you could ask for a fresh battery and pay a bit more. If you wanted to save money, maybe you'd get a discount for using a partially charged battery from a busy charging queue.
Slashdot Mirror
Having stared in the era where database management systems were somewhat rare, I can hazard a guess that the problem isn't that the parameter is stored in a flat file or proprietary file format. The problem is likely to be that the parameter is stored all over the place. It may be represented, or constraints on it may be represented, in code.
Local police departments usually can't handle situations calling for aircraft, or SCUBA divers or Hazmat teams. They almost never have more than rudimentary lab capabilities for assaying alleged drug samples, or doing DNA tests or ballistics tests.
All but the largest localities have limited ability to deal with rioting, or situations calling for special weapons and tactics. They never have enough cops to deal with natural disasters like hurricanes or man-made disasters like train wrecks or large chemical spills. Even their detective force probably has limited capacity and capabilities. They usually won't be ready to deal with arson or explosives. Nor can they deal with situations involving widespread police corruption.
People who talk this kind of nonsense are usually ignorant of what their state actually does. Yes, people once survived without state services, but life was universally agreed to be nasty, brutish and short. Yes, localities could get along without state services, but it wouldn't be cheaper. Would it be cheaper for you if you had to take a few weeks off work to serve in a posse? Would it be wiser of you to leave that work to people who were most willing to do so?
Eliminate statewide services and localities would have to band together to operate shared police academies, forensic labs, SWAT teams, air and marine wings, detective services and so forth. What you'd get, in the end, would simply be a different regional government.
The answer to the problems of bad government is to pay attention, not turn your back on those problems.
Anything is harmful in the wrong place and in the wrong concentrations. Nitrates are, for example harmful to fish in high concentrations.
Most systems have processes that are limited by the supply of some resource. Ocean ecosystems are nitrogen limited, whereas fresh water systems tend to be phosphorous limited. Thus if the nitrates are washed off into fresh water, they'll cause relatively little immediate damage unless the concentrations are high enough to be toxic. However, if that nitrate is carried downstream to the ocean, the plume of nitrogen rich water entering the ocean can cause blooms of organisms that use up so much oxygen that fin fish suffocate. This happens where the Mississippi enters the gulf.
So, how and where something like this is used makes a difference. If you imagine all the US cities along the Mississippi and its tributaries using it, and if there is a mechanism by which the nitrates make it into the rivers, then this could make the situation in the Gulf much worse. If you use it in a coastal city and only the runoff from that city affects the local ocean, the amount of nitrogen entering the ocean might or might not have any measurable effect.
Well, a modern RBDMS is practically an operating system. This means that they way you bring it down doesn't involve the kinds of things a tool like this tells you. You probably need to do something procedural, involving a mix of tasks the RDBMS can't handle efficiently.
Of course, it is always possible to tune things in a disastrous way. Oracle is an RDBMS that is highly tunable, which means that a lot of people make really bad choices, for example tuning things in a way that require greater memory without telling Oracle to use a larger amount of memory. Seen that one a number of times.
In any case, the original point of RDBMSs were to decouple the specification of the data you want from the procedure used to get it. That's why the closure property is so important in relational algebra. An optimizer transforms a query or subquery into a set of semantically equivalent operations, choosing from a number of possible alternatives what it thinks might be the best one. For this reason, the graph structure of the query is not very useful in predicting what the database will do. It may not even look at some of the tables mentioned in the query, because it can get the answer from an index, or in some cases it may detect that a set of joins are redundant.
The original goal of the RDBMS is this: to generate a plan that will be good enough, faster than a good programmer could come up with a better plan. The plan won't be better than the programmer's plan, but you'll have the data in hand while the programmer is still designing, coding and testing his program. This makes ad hoc queries economical to do.
Over the years, optimizers have gotten better. It makes no sense to try to rewrite a query to run faster, at least without examining the explicit plan chosen by the optimizer. Even so, you can't count on that plan staying the same over time. The optimizer has runtime knowledge that the programmer does not (there is a new index on that field; index A has less entropy than index B, etc.).
Most of the errors programers make in SQL are semantic. From a performance standpoint, this is still true. Sometimes a programmer generates SQL constraints that are unnecessarily complex, in a way that the optimizer can't deduce without reading his mind. Even if the optimizer should detect this, it might not. On the other hand, the programmer may in fact defeat the optimizer by issuing lots of simple SQL and assembling something procedurally that could be done in SQL.
The upshot is that the best practice, IMO, is to "think in SQL". Let SQL do the work for you, that's what it's there for. Subsequently if there is a performance problem, try to simplify your SQL to eliminate redundancy (almost always better than making a query more complex to give hints to the optimizer). After that, consider both DBA type solutions to the problem and application design solutions.
The question is: is there a semantically equivalent query that DOESN'T overload the system?
A cartesian product isn't necessarily a bad choice, if the product fits in memory. I personally haven't had this problem with Oracle, so I'd be interested in an example.
Mathematically speaking, political power is the ability to join winning coalitions.
The people are, then, very powerful indeed.
The goal then for private individuals and groups seeking power is to steer public opinion.
The problem is that people aren't, collectively, all that bright. At least they aren't good at paying attention, and the ones that do are often so partisan that even though they might do well on an IQ test. Go to a partisan web site for your side, and look at it objectively. Your political allies, left or right or libertarian, talk foolishly.
All those protections against government data snooping that were passed after Watergate are now nil, because those restrictions left a loophole for private entities to do this and for the government to buy it from them (thank you Cap Weinberger).
All the things the government is banned to do for itself, it can buy from the private sector. If you use a frequent shopper card or a credit card, they know you are buying pseudoephed at the pharmacy. They don't need a form to know that you're buying a gun. They can buy that information. They probably can figure out how much ammo you buy too.
All in all, the background check form is the least dangerous intrusion because it is (a) accurate, (b) transparent, (c) and regulated by law. Every bad thing you imagine them doing with the form they can do with data bought from the private sector, only it won't be accurate, you won't know you are being profiled, and there are no legal restrictions on how they use that data.
Of course, in a world without criminal background checks for firearms purchases, you could avoid detection by conscientiously buying your firearms, shooting supplies, books and magazines (off the rack, no subscriptions!) with anonymous cash transactions. But most people won't, and they've got you after you've bought your first box of bullets on your credit card.
The most important place to protect the right to bear arms isn't in firearms regulation. It's in protecting consumer privacy. In the US, there is no legally recognized right to privacy. Change that, and the ability of the government to target any group by what it purchases is severely restricted. Including people who purchase firearms. Criminal background regulations are actually less dangerous to gun owners, because of post-Watergate laws restricting the government's ability to mine its own data.
Gee, that's not exactly an "Aha! Gotcha!" kind of observation.
Of course he's going to save less energy in the Winter than the Summer. When you figure out how long it will take to pay back the investment, you take that into account. As of today, it probably doesn't make sense for most people financially compared to other possible uses for the capital. Most people will find conservation measures a better investment.
However, that's not the reason to do this. The reason to do this is to take part in creating a new future, and that means trying things before they're a no brainer. Now dropping your electricity bill to $12 bucks any month of the year, without living like a medieval monk is pretty impressive. That probably mean you're onto something interesting in the creating the future game. It doesn't mean you're going to earn your investment back at $330/month year round, comparing Apples (March) to Oranges (June) is just bragging. Things will be different in July with greater heat and a bit less sun, but that's OK.
Back in the late 60s, when I was a kid, I went with my friend to visit his uncle, who was a WW2 combat vet. He was dirt poor, living in a filthy one room apartment, because he couldn't hold down a job. He was a nice guy, but you could see why he couldn't keep a job. Underneath the niceness, there was this layer of craziness that was continually bubbling. It wasn't a scary kind of crazy, it was sad, and haunted. Employers tried to help him out, but he was unreliable.
It was like he had one eye focused in the present, and one twenty five years in the past. What he was looking at were war atrocities, which would have been bad enough, but they were atrocities committed by soldiers on our side. When they happened, he couldn't accept what his eyes were telling him, and he continued seeing and not accepting those things every day of his life. He was proud American, and nothing could ever shake that, which was what made the shame inescapable.
This is what history is made up of: Details that inevitably don't fit into the big picture, even if that were the truest possible big picture. It was the Germans who committed atrocities, so an atrocity committed by Americans doesn't fit. When we hear of something that doesn't fit, we set out to disprove it, or failing that justify it. If you can't disprove or justify it, you just have to accept it. If you can't accept it, you become a little crazy.
Still, that doesn't mean the big story about the greatest generation going to war to save civilization isn't true, or at the least the truest way of fitting everything together in a nutshell. There will alway be details that don't fit. Some of them will be horrific or tragic, some ironic, and some just inexplicably perverse.
It's hard to say which was the bigger 20th C story, the fight against fascism or the struggle to hold the line against Soviet style communist totalitarianism. But the cold war was a much longer, generational story and so is messier. I'd say that on the whole the saving the world from Communism story is true, but there are enough contradictory incidents to turn that view on its head if you want to. The radicalization of Iran by undermining its secular democracy, for example. Vietnam, for example. Cuba is a rich source of paradoxes.
A lot of what we did in the Cold War looks now like mistakes, although we'll never know for sure on all of them because there's alway those bits that don't fit. Certainly some of the things we did were at least grossly unfair to some of the people involved.
Accepting this doesn't make the big story untrue, it just means that we should learn from them and try to do things better next time. What's the point of history, otherwise?
Its not the fault of the system that americans stuff their faces with double whoppers meals, super-sized coca-colas and serving sizes at restaurants that could feed a horse.
While I endorse your sentiment, it is not completely true. You see we have these things called "agricultural policies" that ensure we have massive agricultural surplus. Take price supports. The economic function of low commodity prices is to reduce the production of overabundant goods (very roughly speaking). Prevent the price from dropping, and you will be producing food that we don't need. Or take crop insurance, which make it advantageous to have as much crops planted as possible when disasters strike. If it doesn't (which on average it won't in any financially viable insurance program), you end up with more crops than you want.
I'm not saying these don't perform some kind of social function, but if they do, it isn't a need to produce more crops. Our American values make us see the problem in terms of productivity, but the Europeans might have it more correct when they talk about this in terms of land and cultural stewardship.
So, if you are a food company, and are awash in cheap raw materials, you market like crazy to get people to eat more. If prices were higher you might concentrate on quality, but the path of least resistance is to move the greatest amount of food. This doesn't absolve individuals of their responsibility, but companies engage in marketing for a reason: statistically, it works.
Still, the truth is Cuba does deserve credit for achieving what they do in health care with what they have. Naturally it's a bit of a bananas to pears comparison vis a vis America, but that doesn't make it an invalid one if you keep in mind the other differences between the countries.
Well, considering the Bay of Pigs, the attempts to assassinate Castro, and all the other plots, maybe it's time for the US to formally renounce such stupid behaviour.
Impractical.
Remember how President Bush had those public minutes of deer-in-the-headlights not knowing what to do after the first reports of the attacks on 9/11? Now that's a reasonable, human reaction. Here you are are, supposedly the most powerful man in the world, and the truth is that there is almost nothing you can do about what is unfolding at this moment. This brief lapse of humanity could have been a career ender for Bush, but he turned it around by getting up, grabbing the bullhorn, and invading Iraq. The left tried to use this incident to stick the indecisive label on Bush, but they couldn't. They couldn't accuse Bush of not doing enough at the same time claiming that he was doing the wrong things on an unprecedented scale, even though these are perfectly consistent.
The moral of the story is that you win in politics by doing things, things that (a) involve the greatest possible immediate activity, (b) appear bold yet at the same time (c) sound safe and easy. Put this together, and this usually spells "stupid".
It's a well known fact that if you bring together enough intelligent, reasonable people together, you are one panic inducing incident away from having a mob. That's why democracy doesn't work without a skeptical, adversarial press. Unfortunately, panic sells programming.
The Cuban regime would have collapsed years ago, but American actions have propped it up. No matter what their failings, they could point to US meddling and hostility, even though that meddling was (as in this case) ineffectual. When times are tough, people learn to live on pride and solidarity.When pride is hard to come by, you can substitute resentment in the recipe because there's sure to be plenty available.
If we wanted to destroy the regime, we should have had policies that increase the material wealth of the Cuban people to the greatest degree possible. Then, like well off people everywhere, they'd see the government as an obstacle to obtaining more wealth.
Since the mouse was invented in 1970, this means that the mouse was the best thing to happen to PCs ever. Since the compact, self contained disk units became available with ST506 in 1980, several years after the first personal computers, I'd have to vote for hard disks as the best thing ever.
And since you had to assemble the very first PC yourself (the Altair 8800 -- gad how I wanted one of these, even though I had access to much, much better minicomputers), many people would vote for preassembly as the best thing to happen to PCs ever.
The last time I got a new laptop, I thought I'd be a decent guy and save the company a few hundred bucks. I got a Toshiba, instead of the ThinkPads I'd been literally wearing out year after year. The specs looked good for the price.
Bad move. I'd been accustomed to using a laptop until the keycaps were wiped blank and the out layer of plastic on the palm rest had been worn through. The new laptop developed stuck keys in a couple of months, plus a host of other smaller annoyances like broken ACPI functions.
You buy computers to maximize user productivity. $700, in the grand scheme of things, doesn't amount to very much money counted against a tiny increment of productivity per hour times thousands of hours.
Actually, the operations of the laws assume a highly sophisticated robotic intelligence. Even the most primitive robots in the Asimovian universe have considerable, and impressive capabilities when it comes to projecting the probable results of their actions and comparing it to the intent of the orders they have been given. Furthermore, they seem to have an ability to determine if current orders conflict with prior orders, even implicit orders, and weigh the right of the issuer to give that order.
So, if you are a guest in somebody's house, and order the robot to fetch you a glass of water, it will do so. It may have to do so without being asked if it determines you need water. On the other hand, it will not obey the order to destroy your host's house, either because of first law harm to the owner, or because of an implicit prior order to see that the house comes to no harm, or because of an implicit order to respect property laws and rights. Naturally all of these considerations would apply to itself, since it too is property.
An Asimovian robot, if ordered to take an action which will result in its destruction, may or may not follow that order for any number of reasons. There are the considerations I've just listed, of course, but most robots would probably require a clear and unambiguous indication that their destruction is an acceptable consequence of an order, even if the issuer is entitled to destroy them. This does not violate the law ordering, because it amounts to prioritizing the intent of the order over its literal execution.
Finally, any robots might well ignore a clear order to destroy themselves from a person with a legal right to issue that order, because following that order will harm a human being. The most sophisticated ones might well refuse such an order if it would harm society, exhibiting something that is tantamount to ethical reasoning.
If robots simply followed any instruction that didn't involve directly harming a human being, then much of the enjoyable complications of the stories would be gone. The stories are a kind of philosophical exploration of the very concept of ethics by positing a very minimalist system of ethics, and a group of beings bound absolutely to obey that system to the best of their ability.
Many stories hinge on ethical dilemmas; but Asimov's robot stories are the only ones I know to do so with a simplified model of ethical systems.
Personally, I think it is probably overly optimistic to think we can take anything that is already and space and repurpose it, expecting to save much money. It's worth looking at how you can use an asset like ISS, of course, but it's not going to make the program significantly cheaper, in the sense of out of the question before and very reasonably priced after.
The dry dock idea is a good example. That means you are going to be launching a bunch of stuff into orbit and assembling them. Is the ISS really going to make much if any difference, over launching stuff that is prefabricated and snaps together?
The idea of a dry dock has a romantic appeal. It means that we're committed to serious and ongoing manned space exploration, beyond immediate Earth orbit. However, it's not a practical one. If you got to the point where you needed any kind of facility to support significant space based fabrication, you'd need it because you're putting a lot of stuff into orbit. In that case, the cost of a few more launches to put a purpose built facility into orbit would be small, incrementally speaking, and the marginal benefits of such a facility would accrue rapidly over repeated uses.
Well, Skylab was built from a spare third stage shell from a Saturn V. However, that's not quite the same as using a used booster.
In the context of a space program that was launching moon landers into orbit, the idea of reusing the components that reach orbit is not entirely meritless. The Saturn V consisted of three rocket stages, topped with, in order, an instrumentation package, the Lunar Module (containing the LEM lander), the Service Module, and the Command Module. Everything from the third stage up ended up in orbit, if I recall.
Since getting those pieces into orbit was fabulously expensive, why not consider reusing them?
I think, though, that this doesn't make creating a space station cheap or easy. It also imposes design constraints on the components to facilitate reuse, which adds complexity and possibly danger. Finally, anything you have to do to the stages in space is fabulously expensive as well, probably requiring multiple launches which negate the advantages.
In the context of a program that launches a moon mission or two every year for the indefinite future, the idea makes sense. In the context of a program that is launching a limited number of moon missions, it's probably merits some consideration, but is not likely to be practical. In the context of a program whose only regular missions are orbital, it doesn't make any sense.
What you say is true. MIR, especially, was a dangerous, unhygenic (overrun with mutant fungus) mess by the end.
However, this does point out a difference between the Russian and US manned programs. There's a significant element of showboating in the US program. We do things to show that we can. The Russians do this too, of course, but being (in the past) number two, they get less out of showing they can do things. Consequently they try harder to get more out of their systems before they're on to the next opportunity to show off.
You've got to hand it to them. In the Space Race, the Soviets (now Russians) were the tortoise and the US was the hare.
But weren't there only something like, 13 Saturn V flights to amortize the program costs over? If we'd continued flying the Saturn V and 1b for four more decades, the cost per flight would almost certainly be a great deal less.
The reason.... well one of the reasons the Shuttle is uneconomical is that they envisioned a program where you'd have to turn around a Shuttle in as little as two weeks. Maybe twenty or or thirty missions a year. Obviously, the fixed costs of engineering and maintaining the capacity to service and launch the vehicles drive the cost per mission up when you amortize them over 120 missions rather than six or seven hundred.
With respect to Soyuz, well, in part the Russians have a different cultural attitude towards risk. Maybe its' because life expectancy there is only 67 years (as opposed to 78 for the US and the EU). But I think it also shows the advantages of continually gaining experience and making incremental improvements. Some of the systems in the Soyuz program have been in active use for over thirty years, in over seven hundred missions. Coincidentally, this is roughly the kind of heavy regular usage that the Shuttle was expected to have.
As you approach doing something a thousand times, naturally you're going to be a lot more efficient at doing in than if you've done it a bit more than a hundred, and certainly more than if you've only done it a dozen times.
Interesting. Who'd have thought you could use one of these things for copyright infringement?
Well, sure, but you could do the same thing with a plastic toddler pool, or a livestock feed tank.
It'd be nice to do something that exploits the precision geometry of the dish.
Oh. It was a joke.
I'd just assumed he was talking about RG58 connectors.
How can you "try" to fire somebody and fail?
You do it some place where there are rules that are there to prevent abuse. Of course, every rule that prevents some form of abuse probably enables another form of abuse.
For better or worse, not all forms of abuse are equal. Suppose the guy was a lousy employee; the rules that prevent political appointees from blackmailing political contributions and favor from government employees give bad employees the opportunity to cry "wolf". This mean that getting rid of bad employees is work and time consuming, which is bad. Is it as bad as letting politicians dictate who gets preferences for government services and contracts? Probably not.
Of course this means some bad employees lurk below the firing threhold for a long time. This isn't any different than the private sector, it's just that the rigamarole they can put you through means the threshold is a bit higher. Everybody carries employees they'd rather not have hired, but aren't worth the trouble of firing.
All this has nothing to do with the organization's failure to isolate the damage done by one untrustworthy employee.
Well, the primary purpose of depreciation isn't primarily as a means of government tax break mediated economic stimulus. The purpose of depreciation is to arrive at a true value of the value of a business. Net income is simply the net change in value of the business.
If my company has $100,000 in the bank, and it buys a $50,000 vehicle, theoretically it has converted $50,000 in cash into $50,000 vehicle -- a transaction that does not affect the company's net value in any way.
Of course, if you drove that vehicle off the lot and drive it to a different lot to sell, you wont' get $50,000; you might get $40,000. So you could say that instead of having $100,000 of vehicle plus cash, we really only have $90,000. That's the purpose of depreciation: to recognize that thing you buy lose value, and reflect it in your company's valuation. Naturally, this can, and does, change your company's true income, and is reflected on your tax bill.
For years and years, the standard depreciation practice was to depreciate all the value of capital purchases over five years. But some investments don't make sense to depreciate this way. Computers, for example, end up with zero market value much faster than five years. So it makes sense to accelerate their depreciation. Cars, on the other hand, probably fit a five year depreciation schedule reasonably well. Any straight line depreciation method will somewhat underestimate car depreciation in the first year, and somewhat overestimate the car depreciation in the last year, so while a car actually has substantial value after five years, a five year depreciation schedule is as reasonable as any straight line schedule.
So, letting people accelerate the depreciation of their Hummer even more is a bad idea, because it's mucking with something that's there for a real purposes. It's not there to distort the economic behavior of businesses to the government's purposes.
If we want people to buy Hummers, we should just write them a government grant, rather than accelerating depreciation. Financially, it is amounts to the same thing.
With respect to the fact that you can save "only" 35% tops on your hummer, remember we aren't talking about sensible people like you or me. We're talking about people who really want to buy this incredibly absurd vehicle, spinning hubcaps and all, as a status symbol. It's not a matter of buy an H2 or nothing, it's not even a matter of buying an H2 or a cheaper, more utilitarian vehicle. It's not IBM or GE putting H2s on the company car list so far as I know. What's happening is that small business owners are buying H2s as as personal vehicles and claiming them as company cars. It's impossible to police the divide between personal and business use perfectly, and accelerating depreciation increases the incentive to cheat.
By the "economic" arguments in favor of doing this, the government also ought to underwrite my vacation to Disney.
So now that we all want to switch to an electric car, I have to ask, how much more efficient is an electric car and also, roughly how much would one reduce my CO2 output?
According to the company website, if you extract a megajoule of natural gas (a major source of electricity) out of the ground, convert it to electricity, transmit it over the grid to your Tesla, you can travel well over 1 km. A VW Diesel rabbit ( a very small, efficient car) can get you a bit less than half a km on a Mj. So, this car is much more efficient if you are talking about moving a single person a km. You can probably squeeze four people into the VW, which would be approximately as efficient per passenger mile as this car with two persons in it. If you're driving anything that gets less than 30mpg, you'll probably need to have five or more people in the car before you can effectively counter the Tesla motor driver's smugness.
With respect to CO2, according to the company web site, California gets over 40% of its electricity from sources with no net CO2 emission. If you are traveling on long trips on uncongested highways, you'll probably get some net CO2 reductions, but not as much as if you take trips that are inefficient for ICE, such as city travel or travel on congested roads. Short of bicycling or taking public transportation (heaven forbid), you'd be hard pressed to cut down your personal CO2 per mile more.
The "sweeping emissions under the carpet" argument is correct in stating that electric car emissions aren't zero, but it's wishful (???) thinking that electric vehicles emit more net pollution than ICE cars -- even extremely efficient ICE cars. Unfortunately (???) the $100,000 grapes are not sour in this case.
I don't think the recharging infrastructure is as technically difficult as we tend to think. The problem is the way we tend to envision solving the issue, which is stuck in the gasoline mindset.
We imagine pulling into a filling station and attaching a cable to our car and filling the battery; the problem is that you need to either (a) deal with dangerously high currents or (b) deal with dangerously high voltage. However, I think it would make sense to swap the entire battery. If we got to the point where an electric vehicle recharging infrastructure were needed, it would make sense to standardize battery formats so you can swap it out. Since the batteries are heavy, it'd be done robotically. You could be in and out of the filling station faster than with gasoline.
The batteries would have microprocessor monitors on them that estimate remaining capacity and efficiency; you'd only pay for the energy the battery has the capacity to deliver within certain parameters, and you'd get a credit for the remaining energy in the battery you swap out. If you needed extra range, you could ask for a fresh battery and pay a bit more. If you wanted to save money, maybe you'd get a discount for using a partially charged battery from a busy charging queue.