If some guy, being a mouth-breathing moron with downs, slam your car into the back of their car at 60mph, that's 100% your fault. The insurance company will rightly blame you 100% for hitting them.
Trying again...
If some guy, being a mouth-breathing moron with downs, slam your car into the back of their car at 60mph, that's 100% his fault. The insurance company will rightly blame him 100% for hitting you.
Keeping context straight is hard sometimes. But I'm running on enough stim to not really need sleep so this is just blatant inattentiveness... like driving down the street and hitting a parked car.
No, not really. They share responsibility for "obstructing traffic", i.e. for being inconvenient. That's it. That's 100% of their responsibility.
If some guy, being a mouth-breathing moron with downs, slam your car into the back of their car at 60mph, that's 100% your fault. The insurance company will rightly blame you 100% for hitting them. That's not an "It's you or him" insurance thing; that's a legal thing. Legally, you, being hit, are responsible for exactly as much as you are responsible for not being hit. He, being the guy who failed to fucking control his moving two-tonne machine in a way that would have allowed him to operate safely in reasonably likely scenarios, is wholly responsible for his improper control of his car.
You see, the law says you are supposed to be in control of your vehicle. A woman sitting stopped at a green traffic signal texting on her phone IS IN CONTROL OF HER VEHICLE. Her vehicle is stopped. It is, currently, parked illegally and obstructing traffic; but she is in control of it. An idiot ramming into a parked vehicle in the road way just failed to control his moving vehicle and collided with what was not but could have been an ambulance, fire truck, police cruiser, disabled vehicle, pedestrian, slow-moving vehicle (tractor, EPMAD, bicycle), confused deer, construction worker, or whatnot. Only one of these people actually failed to control their vehicle; one was in perfect control of her vehicle, but was being a dick.
Put simply: My vehicle is under control; you can complain that I have it under control in a manner you don't like and is unlawful, but you can't assign responsibility to me for your vehicle not being under control. You aren't responsible for vehicles that operate in a manner that puts you at unmitigatable risk--a vehicle that suddenly pulls out in front of you forces a faster than reasonable reaction and may leave you with no ability to avoid or brake, and thus is at-fault--but you are legally and ethically responsible for recognizing stationary objects and being fully capable of avoiding or braking.
It would be like ramming a shopping cart into an old woman standing in the middle of the aisle instead of closer to one shelf so others can pass, and then blaming her because she shouldn't have been there. Why don't YOU watch where you're fucking going with it? She's inconsiderate; you're a danger to those around you, an asshole, and a moron.
As I was comparatively defending MongoDB, which was facing criticism for being "unreliable" compared to traditional SQL-interface fixed-column row-based relational database software, I felt the need to point at the documentation for PostgreSQL (a top tier product) while explaining the potential for data loss in its replication configurations. Trying to spout a bunch of fluff about how PostgreSQL typically has a window of data loss in replication configurations aside from an expensive synchronous setting without actually bringing documentation to back it up could easily be interpreted as selective reporting and poisoning the well.
I didn't feel the need to cite MongoDB's docs because they are incredibly easy to dig through (PostgreSQL's docs take some extra legwork to find something relevant--google can hit it if you know what you're looking for, but there is a lot of WAL and replication stuff that doesn't quite contain the wording and explanation I need; the page I found is perfect for this) and because it's harder to apply bad faith to that part of the conversation. To discuss PostgreSQL in bad faith, I would have to omit things like PostgreSQL WAL synchronous streaming; to discuss MongoDB in bad faith I would have to insert false statements which can be quickly verified as false.
I will say this again: Throw out that stupid outdated "United States Constitution" and write a new one. We're way too far into broken hacks on top of hacks in this alpha-quality code to keep running it in production. One of the very large and obvious defects in this code is the complete failure to implement a good process for creating laws.
Every law should have a simple, plain-language English scope statement written at the top as a preamble. Every single bill. That states the goals of the law and the scope. If technically possible, the scope statement should be written originally in Latin and/or Ancient Greek; it is acceptable to write part of it in Latin and/or ancient Greek to solidly clarify certain descriptive portions while referring to the English portions of the law that specify technical concepts too new to translate into Latin and/or Ancient Greek properly.
An English scope statement will leave you with stupid shit like "A well regulated Militia, being necessary to the security of a free State, the right of the people to keep and bear Arms, shall not be infringed" ending in arguments on whether "the right of the people to keep and bear Arms" references the National Guard or the actual individual citizens. A large part of that argument is over whether or not the meaning in 1780 was different from the way we'd interpret this sentence today. That is: a sane, rational, and learned person might generally interpret this to reference the People as a body and thus allow them to have their state raise a military force not under jurisdiction of the Federal Government; however, a sane, rational, and learned person *in* *1780* might correctly read this as that every individual person has the legal right to carry any weapon he wishes.
Latin and Ancient Greek are well-understood dead languages. Sanskrit is also well-understood and dead, but less generally accessible: as little as 50 years ago, Latin and Ancient Greek were both standard core education in America; they still are in some public and many private school systems. A firm working grasp of both Latin and Ancient Greek were necessary to enter top-tier colleges several decades ago. Statements written in these languages have exactly one interpretable meaning, or can be made to have exactly one interpretable meaning; that meaning won't change with linguistic drift over time, as with English, and we won't have to speculate over whether or not a sane, rational person in 1780 meant that each individual should have guns or that the people should be able to raise a non-Federal military force because the specific meaning is right there in clear Greek.
As for actual laws, it should be a 100% bullet-proof legal defense that the law is invalid because it doesn't accomplish its scope. The scope says the law attempts to implement certain measures and for a certain purpose; if either the measure implemented in the part of the law you broke isn't in the scope statement *or* the action you took doesn't fit with the reasoning behind the law (for example: operating your cellphone GPS while parked at a light does NOT pose a danger to others through distracted driving increasing the risk of mis-operating a moving vehicle), the law is not applicable and has no standing.
Remember: The law must satisfy the scope as explained in Latin and Ancient Greek. Any conflict in the English scope statement with these is overruled by these. The Latin and Ancient Greek must not conflict with one another. The Latin and Ancient Greek may be partial clarifications and explanations of the English scope statement, explicitly yielding specific parts of the English scope statement to interpretation. They may even cite that a certain part of the English scope statement cannot be sufficiently explained in Latin or Greek, put bounds on the scope, offer guidance, and yield remaining interpretation as does not run afoul of these bounds and guidance. The full text of the law is in English, but is not valid outside the bounds of this scope.
The economic cost of a minor traffic jam caused by a 5 second delay in moving your stopped vehicle and NOT getting a ticket because the police made a judgment call that you're not doing sufficient harm is lower than the economic cost of the PD processing a ticket, the driver having to go to court and take up the court's time, the driver having to expend several hours of their time in travel and time in court, records having to be expunged, etc.
It's also, amusingly, lower economic cost than fining the driver and/or pointing their license--which means, in the strictest sense, there is a net societal gain from not issuing citations like an autistic retard to drivers parked at a traffic signal using their phones!
Me and my friends actually devised a game where we go speeding along the highway at night (the only time speed cameras are active) and jumping red lights on open, clear intersections where you can verify it's safe. The cars have high-power dipole canons mounted and we try to EMP the cameras just as we get close. The more daring you feel, the closer you get before reducing the electronics to a smoldering heap.
Actually it does matter. I have had cars stall and fail when attempting to accelerate from a stop. Discovered the spark plugs got fouled because of a blocked EGR valve. The car ran fine until that moment.
It is 100% the responsibility of the driver who collided with the obstruction for colliding with a non-moving obstruction. The person sitting at a light could be cited for obstruction of traffic; however that does not move the blame for the collision onto him. A person sitting at a light being a dufus could easily be a disabled car, a cop car, a deer, a construction worker, a driver experiencing a heart attack, or any number of things that the next driver needs to remain aware of and react appropriately to.
Or are you the type that also blames gun manufacturers for murders?
The job of a police officer is to ensure the smooth and peaceful running of society. Police officers are supposed to use their better judgment to decide if they need to intervene in an unsafe situation. They're not supposed to be walking porcine bureaucrats looking to randomly drain and damage society by mindlessly misapplying draconian rules into incidental situations.
Texting while driving laws were put into place because of the extreme danger of distracted driving. That danger isn't present when parked at a light--you might annoy someone by not moving when the light changes, and you'll obstruct traffic in a non-dangerous manner. We have accepted the danger of people referencing, but not programming, their GPS while driving; we certainly haven't targetted GPS use while parked at a light. Ticketing people for these things is inappropriate, regardless of what the law actually says. The law was put in place specifically to address certain societal problems; these actions do not intersect with those problems, and so the officer should apply his legal discretion rather than acting like a predatory dickhead.
PostgreSQL has settings that make up that last bit of reliability, putting it roughly on par with MongoDB.
And...
Short version: MongoDB allows you to, on a per-query basis, write data into the database at any level of reliability that MySQL and PostgreSQL provide. Single-server Journaled (WAL log shipping, WAL asynchronous streaming, MySQL master-slave replication), multi-server Replica Acknowledged (PostgreSQL WAL synchronous streaming), and a single-server "Acknowledged" mode that is faster but gives a weaker data durability guarantee (transaction is valid, not yet to disk, and not replicated).
"Extremely unlikely" comes out because the data's been replicated enough that it should be propagating further. Notably, 50%+1 nodes, so 2 nodes in a 3 node cluster, or 3 of 5, or 5 of 7. The data is also flushed to disk every 100mS ("j" write concern on a single node flushes the journal immediately and won't acknowledge until it's on disk). Once the data is to disk, it's 100% guaranteed to be there if the server experiences a recoverable system failure (power drop, kernel panic); it will be lost if it experiences filesystem corruption or a hard disk failure.
If you have more than 50% of the replica set go down, the whole cluster stops. The cluster is now refusing writes. If a replica node comes up with the data on disk, it will successfully propagate through the replica set and the data will not be lost.
To fail this, you need every replica node to acknowledge that it's got the data and then immediately fail in the 100mS window between acknowledging the query and writing it to disk; or you need them to suffer fatal hard disk failure or file system corruption and a system halt before they forward the data on to the others in the replica set. All of them. If you have a set of three, you need the primary and one of the secondaries to both fail; the primary won't acknowledge the transaction at all until 50%+1 of the replica set has it (one secondary in a replica set of 3), so the window of failure there is actually quite small,
As you can imagine, this involves a very carefully orchestrated exact sequence of catastrophic events each with low probability of occurring over long timescales. A system failure in any given 8.76 hour period has a 1 in 1,000 chance in a system that has 99.9% uptime. We're talking about two systems failing in exact windows of 0.1 seconds in tandem; two systems failing in the same 8.76 hour period has a probability p=1/1,000,000 with 99.9% uptime. There are 315,569.26 0.1 second periods in each 8.76 hour period, giving a probability of failure in any of these at the exact moment required of 1 in 315,569,260; probability of two exact failures in the correct way for a replica set of three is 1 in 99,583,957,900. With 99.99% uptime, it's 1 in 99,583,957,900,000,000. Note that the window is actually smaller than 0.1 seconds because it'll occur between a journal flush most likely, so it's more like a 50mS window on average and the numbers I gave should be doubled (1 in 2n).
It's a small window. PostgreSQL's synchronous WAL replication is 100% guaranteed if it acknowledges the transaction, because it doesn't acknowledge until both servers hit the disk; the only data loss possible is the even less likely case of both servers destroying their disks completely at the same time (actually this is probably more likely: the window of data loss would be, at a minimum, the time it takes to react to a failover by running an immediate automated backup; and even then, you're no longer streaming new data to another server). In a three-way replica set on PostgreSQL with synchronous WAL (Master and two Slaves), you need three such failures and thus your risk of data loss is slightly lower than MongoDB (again, because that third server will 100% guarantee data to disk before you get a successful COMMIT; but then you've got a wider window for disk failure on the last one).
No, I mean MongoDB will take a 3 database cluster and let you "Replica Acknowledge" a transaction with "Majority" count. Once it hits 50%+1 servers, it's 100% guaranteed solid unless you lose both servers. If both servers suffer a power drop at that point, the last server refuses to accept writes; when those servers come back, they will replay their oplog back to the last server to synchronize it. There's one flaw here: there's no "Replica Journal Acknowledge", so it's theoretically possible to lose that transaction anyway; both servers have to suffer a system failure (power drop, kernel panic) within 100mS of receiving the operation, since they write out their data to disk every 100mS. In practice this is extremely unlikely.
That means once you've sent it and gotten back that it's written, it is written. You'd have to lose both (or more--3 servers in a cluster of 5, etc.) servers' power or hard drives (corruption, failure) before the data is propagated further.
By contrast, Percona and MariaDB have XtraDB. XtraDB does optimistic locking: in normal autocommit, the transaction might get rolled back silently--it will write successfully to one server and return success, but if another server simultaneously gets a write that conflicts and starts propagating it then the transaction will be silently rolled back (i.e. undone, removed, lost, failed). With BEGIN-COMMIT transactions, you may get a Deadlock on "COMMIT" and then you're informed that it did in fact roll back the transaction and you must re-submit (i.e. do this if you actually care about durability of the data). With autocommit, as well as with any transactions (even explicit COMMIT) on MySQL master-slave replication or PostgreSQL WAL replication, you may in fact be informed that the transaction is 100% committed and then have that server FAIL and the slave comes up without that transaction--unavoidable silent data loss.
The failure mode expressed by MySQL master-slave replication and PostgreSQL WAL replication in the default asynchronous streaming replication mode is the same failure mode as with "Journaled" write concern in MongoDB. When running "Journaled" rather than "Replica Acknowledged," you write to exactly one server and are told it's committed when it's written to disk--it's durable on that server, but not necessarily replicated. If that server power drops and comes back up, it may find new operations have made its non-replicated operations invalid; it will then silently roll those back.
Therefor, in cluster layouts, it is possible for MongoDB to have a negligible reliability advantage over PostgreSQL's most common replication methods. PostgreSQL has settings that make up that last bit of reliability, putting it roughly on par with MongoDB. MongoDB has a guaranteed "It has reached enough servers that it is valid on the cluster unless God hates you" write concern by which the data is likely to actually be there if it tells you it's there, unless a very specific subset of servers experience a catastrophic failure in an extremely small (tenths of a second) window--a subset large enough to take down your entire cluster.
Short version: MongoDB allows you to, on a per-query basis, write data into the database at any level of reliability that MySQL and PostgreSQL provide. Single-server Journaled (WAL log shipping, WAL asynchronous streaming, MySQL master-slave replication), multi-server Replica Acknowledged (PostgreSQL WAL synchronous streaming), and a single-server "Acknowledged" mode that is faster but gives a weaker data durability guarantee (transaction is valid, not yet to disk, and not replicated).
*"PostgreSQL streaming replication is asynchronous by default. If the primary server crashes then some transactions that were committed may not have been replicated to the standby server, causing data loss. The amount of data loss is proportional to the replication delay at the time of failover."
MySQL is a piece of shit. MongoDB and PostgreSQL, depending on the data layout (MongoDB if you're looking for an indexed XML-YAML-JSON type of data with reliable data integrity in a cluster; PostgreSQL if you're looking for an indexed CSV with good replication and only minimal probability of basically silent data loss in a fail-over scenario).
These are not poor countries. "Poor countries with the greatest natural resources wind up with the most corrupt regimes" implies that the country started poor and gained an evil tyrant, not that it's poor because of the evil tyrant.
The whole statement is ridiculous doublethink. "Poor countries with the greatest natural resources"? In Burkina Faso, you can get gold by sifting tiny, tiny flecks out of dirt. That's not super-rich great natural resources; there's gold in the dirt and it takes a ridiculous amount of effort to get to it, so that's essentially "resource poor".
If these countries had great natural resources, they would be rich as living fuck. Don't tell me that ubiquitous presence of trace elements means "great natural resources", because that's like saying the bits of water you can squeeze from plants in the desert count as "well-hydrated marsh region." Hell, the desert's better off: you can squeeze a cactus. Imagine that water being distributed evenly across the desert as moisture in soil 6 inches under the sand. Water like a raging river, but you have to acquire it by concentrating what is an unending puff of dampness stretched across the vast and endless desert. More water than in the Mississippi, but at least we could stick a bucket into the Mississippi and get something vaguely drinkable.
They're poor for a reason. It's the exact opposite of having great natural resources.
No, the issue is that a wind turbine produces a lot less than a single, large nuclear power plant. Nuke plants produce several gigawatts if they like; a single wind turbine produces 1-3MW, so if you have 2,500 wind turbines you can kind of approach what one fairly large nuke plant produces. Alta Wind Energy Center has 490 turbines and produces 1.3GW. Kashiwazaki-Kariwa produces 8GW. The US has 90 gigawatts of nuclear power (790TWh produced in 2011) supplied by about 100 small and large nuclear plants.
People are dumb enough to think nuclear power is dangerous. There was a huge news story about a nuclear plant leaking radiation into the water table a few years ago. You know how much radiation it leaked? Enough over 10 years that if you extrapolated the leak by 10,000 years and concentrated that into one acute dose, you might have some long-term health effects a few decades down the line.
Radiation is a big scary boogieman. It's like fire. Oh my god there's a fire! Well if you stick your face into the charcoal grill, you might have some trouble; but overall it's.. warm, a few feet away... and if you're any appreciable distance off it's not so warm. That warmth won't hurt you. If the thing tips over and sets fire to your house and you're entrapped in the blaze, that's different.
People imagine that "it's warm downwind because a concentrated column of air is blowing my way" is the same as "HELP I'M ON FIRE!!!" They get a little radiation leak, or even a failed plant, and they're like... oh god, contamination, contamination everywhere! But the truth is nothing changes, the minimal amount of radiation exposure is largely untroubling, and everything is fine. Unless it's Tjernobyl spewing liquefied toxic waste everywhere and dumping raw liquid nuclear fuel into the water table and scattering tons and tons and tons of radioactive matter into the air, it's a non-starter.
I mean for god's sake, people in Taiwan living in highly radioactive buildings (1000 times background radiation) show 40% lower incidences of cancer and overall have better health; most studies only manage to conclude that we can't decide if the reduced incidence of ALL TYPES OF CANCER (not Simpson's Paradox; all means each individual type--less leukemia, less lung cancer, less skin cancer, etc.) is because the population is younger or because these are mostly rich upper-middle-class people who lead healthier lifestyles. That's a choice between "no real change" and "These people don't eat garbage and have heard about something called 'jogging'".
A radiation plant with a leak will not expose the local population to 1000 times background radiation. It won't expose them to a short-term 1000x dose or a lifetime of 1000x background radiation. Even a major nuclear disaster, unless the plant fucking explodes, probably won't expose the local population to doses like that--what do you think containment buildings are for? In most cases, a leak that raises alarms and exceeds regulatory guidelines and gets the NRA on their asses will expose the population 100 feet away from the leak to less background radiation than living within 10 miles of a coal plant does. And even then, sometimes it's fucking harmless (all this bullshit over a minor TRITIUM leak?!).
Big bogeyman. Nuclear plants aren't as scary as people want you to believe. Look at the Fukushima disaster--people are talking about the disaster, but not about the actual impact. We've arbitrarily assigned a high impact to this disaster, but is there any real impact at all? Sure. I bet some people get cancer earlier than they would normally. Eventually. We'll see in 30 years.
That's rich. Flooding an area has "minimal environmental impact". It only ever really worked at Niagra Falls, since you didn't need to create a giant fucking reservoir where there was a wilderness.
I bet the people of Fukushima suffer minimal hardship. Like, a third of them get cancer, when they're fucking old, and probably would have cancer anyway.
Argumentum ad absurdum is useful when you're trying to make a point. In some cases it is patently absurd to "stick to principles".
In some places in Africa they have effective slave labor. People are "free", but children as young as 5 or 6 work in mines with no safety equipment laboring to dig up little flecks of gold. They develop lifelong health issues. Many die. They do this because resources are so scarce that your whole family right down to your young children must work in such conditions to make enough money to spend every day hungry, but alive with just enough food to keep you sickly and breathing. This isn't just a small segment of the population; entire regions operate like this, with just about everyone scraping by. Tanzania, Burkina Faso, places like that. Merchants show up where the mines are bearing nothing beyond cheap food and tents to sell it from, because these people can't go very far to eat.
Capitalism. Capitalism in a region too poor to support itself.
Capitalism works in America, in Canada (OIL RICH!), in Norway (OIL RICH!), in Japan, in the UK, in France (France has an interesting system that I like, by the way, and it works extremely well for them)... all places where you can sleep in, go to work, and then stop at McDonalds and get a really shitty hamburger.
Capitalism isn't working for some significantly sized regions in Tanzania. Capitalism isn't working for Burkina Faso. And somehow these people can't just leave and go to the adjacent region where they're truly free, and not slaves of necessity. Would that it were so simple to walk five miles to freedom; but apparently five miles is either cripplingly far or there is nothing five miles away any better than what's directly here.
These people would be, to a feudal lord, important human assets. These serfs are valuable. They are not valuable to gold miners because we can always find more; but as an asset that comes with the land, they are valuable to a landowner. When your serfs are sickly, they don't work as well; when they die off, they expend time and energy on dying and death rites instead of working. You must improve these conditions, must find other ways for them to produce, must produce medicine to keep them healthy and technology to make them harvest the resources on your land more effectively to make you richer.
Gold mines are interesting. You have to mine out an area and move to the next before the next guy, so mining at speed is important. On the other hand, larger operations that don't have access to essentially slave labor can't do that (i.e. in the US or Europe); we've developed ways to mine gold by using plants, which takes about 10 years. Now if the landlord owns the land, his only possible way to enrich himself is to acquire more land; he doesn't care if it takes 10 years to mine an area by phytomining (including by digging and then phytomining the pile of dirt, rather than spending a lot of time sifting), because he owns the land he owns. If the serfs can spend their time growing food on non-mine land and building exportable things, living healthier, they're more productive and thus making the landlord richer. Automated mechanical sifters are attractive too.
So take your pick: small children choking on mineral dust, barely surviving to breed, taken as fodder by opportunists in these resource-poor regions; or feudalism, serfs, some semblance of value in the people, and a lot of hardship because resource-poor has a real meaning and we can't just give you a magical solution like "Captialism" or "Feudalism" or "Marxism" and create an instant utopia.
By the by, an interesting thought about Burkina Faso: The whole of Europe and the New World are trying to boycott any gold from Burkina Faso, desperately trying to trace gold sources back to avoid anything from there. They say the conditions there and the child slave labor are intolerable, so they're just going to break the backs of the slave drivers. Once they do that......there
It's cooler than when I read it as "Find millions of nearby assholes" and decided that Craigslist and OKCupid already do this. Or living in Boston or Baltimore, where you just have to look outside.
An enslaved population must survive to have a revolution to gain its freedom back; a free, poor society must survive to develop into a free, prosperous one. Principles are easy to fall back on: just refuse to negotiate with terrorists while they continue walking down the line of hostage 7 year old girls and mouth and anal raping each of them in turn trying to break you.
Non-negotiation with terrorists has real impact. Sometimes people die who didn't need to; but the ineffectiveness of hostage negotiation means fewer people are so inclined. Those that are will raise the stakes, which means that the president's daughter will be kidnapped by ninjas instead of just some random schoolgirl. Although a few suffer today, the many will benefit from their suffering and death: fewer will ever be taken hostage, and thus fewer will be killed as hostages in the future. This is a good place to stand on principle.
Until they're going to nuke seven elementary schools at once.
In a society so poor that "freedom" and "capitalism" results in the extreme suffering and death of 98% of the population--thousands, millions--some control of resources is better. A feudal society is only quasi-slavery: you're effectively "free" in that you can live your life as long as you produce--which is like America, where you're only "free" until you're a lazy hippie who doesn't get a job and tries to smoke pot all day and leech the system that's there for people who are TEMPORARILY down, not permanent willful parasites. In feudalism, obviously you can't go on vacation; you can't leave your land. On the other hand, extremely scarce resources are managed somewhat better via feudalism, so folks get by under appreciable hardship rather than starving and rotting as a society. Sometimes the hardship even gives way to a tolerable, somewhat enjoyable, but not very exciting way of life.
Freedom comes at a price. If you can't afford it, then we take death in exchange.
Slashdot Mirror
If some guy, being a mouth-breathing moron with downs, slam your car into the back of their car at 60mph, that's 100% your fault. The insurance company will rightly blame you 100% for hitting them.
Trying again...
If some guy, being a mouth-breathing moron with downs, slam your car into the back of their car at 60mph, that's 100% his fault. The insurance company will rightly blame him 100% for hitting you.
Keeping context straight is hard sometimes. But I'm running on enough stim to not really need sleep so this is just blatant inattentiveness... like driving down the street and hitting a parked car.
No, not really. They share responsibility for "obstructing traffic", i.e. for being inconvenient. That's it. That's 100% of their responsibility.
If some guy, being a mouth-breathing moron with downs, slam your car into the back of their car at 60mph, that's 100% your fault. The insurance company will rightly blame you 100% for hitting them. That's not an "It's you or him" insurance thing; that's a legal thing. Legally, you, being hit, are responsible for exactly as much as you are responsible for not being hit. He, being the guy who failed to fucking control his moving two-tonne machine in a way that would have allowed him to operate safely in reasonably likely scenarios, is wholly responsible for his improper control of his car.
You see, the law says you are supposed to be in control of your vehicle. A woman sitting stopped at a green traffic signal texting on her phone IS IN CONTROL OF HER VEHICLE. Her vehicle is stopped. It is, currently, parked illegally and obstructing traffic; but she is in control of it. An idiot ramming into a parked vehicle in the road way just failed to control his moving vehicle and collided with what was not but could have been an ambulance, fire truck, police cruiser, disabled vehicle, pedestrian, slow-moving vehicle (tractor, EPMAD, bicycle), confused deer, construction worker, or whatnot. Only one of these people actually failed to control their vehicle; one was in perfect control of her vehicle, but was being a dick.
Put simply: My vehicle is under control; you can complain that I have it under control in a manner you don't like and is unlawful, but you can't assign responsibility to me for your vehicle not being under control. You aren't responsible for vehicles that operate in a manner that puts you at unmitigatable risk--a vehicle that suddenly pulls out in front of you forces a faster than reasonable reaction and may leave you with no ability to avoid or brake, and thus is at-fault--but you are legally and ethically responsible for recognizing stationary objects and being fully capable of avoiding or braking.
It would be like ramming a shopping cart into an old woman standing in the middle of the aisle instead of closer to one shelf so others can pass, and then blaming her because she shouldn't have been there. Why don't YOU watch where you're fucking going with it? She's inconsiderate; you're a danger to those around you, an asshole, and a moron.
As I was comparatively defending MongoDB, which was facing criticism for being "unreliable" compared to traditional SQL-interface fixed-column row-based relational database software, I felt the need to point at the documentation for PostgreSQL (a top tier product) while explaining the potential for data loss in its replication configurations. Trying to spout a bunch of fluff about how PostgreSQL typically has a window of data loss in replication configurations aside from an expensive synchronous setting without actually bringing documentation to back it up could easily be interpreted as selective reporting and poisoning the well.
I didn't feel the need to cite MongoDB's docs because they are incredibly easy to dig through (PostgreSQL's docs take some extra legwork to find something relevant--google can hit it if you know what you're looking for, but there is a lot of WAL and replication stuff that doesn't quite contain the wording and explanation I need; the page I found is perfect for this) and because it's harder to apply bad faith to that part of the conversation. To discuss PostgreSQL in bad faith, I would have to omit things like PostgreSQL WAL synchronous streaming; to discuss MongoDB in bad faith I would have to insert false statements which can be quickly verified as false.
I will say this again: Throw out that stupid outdated "United States Constitution" and write a new one. We're way too far into broken hacks on top of hacks in this alpha-quality code to keep running it in production. One of the very large and obvious defects in this code is the complete failure to implement a good process for creating laws.
Every law should have a simple, plain-language English scope statement written at the top as a preamble. Every single bill. That states the goals of the law and the scope. If technically possible, the scope statement should be written originally in Latin and/or Ancient Greek; it is acceptable to write part of it in Latin and/or ancient Greek to solidly clarify certain descriptive portions while referring to the English portions of the law that specify technical concepts too new to translate into Latin and/or Ancient Greek properly.
An English scope statement will leave you with stupid shit like "A well regulated Militia, being necessary to the security of a free State, the right of the people to keep and bear Arms, shall not be infringed" ending in arguments on whether "the right of the people to keep and bear Arms" references the National Guard or the actual individual citizens. A large part of that argument is over whether or not the meaning in 1780 was different from the way we'd interpret this sentence today. That is: a sane, rational, and learned person might generally interpret this to reference the People as a body and thus allow them to have their state raise a military force not under jurisdiction of the Federal Government; however, a sane, rational, and learned person *in* *1780* might correctly read this as that every individual person has the legal right to carry any weapon he wishes.
Latin and Ancient Greek are well-understood dead languages. Sanskrit is also well-understood and dead, but less generally accessible: as little as 50 years ago, Latin and Ancient Greek were both standard core education in America; they still are in some public and many private school systems. A firm working grasp of both Latin and Ancient Greek were necessary to enter top-tier colleges several decades ago. Statements written in these languages have exactly one interpretable meaning, or can be made to have exactly one interpretable meaning; that meaning won't change with linguistic drift over time, as with English, and we won't have to speculate over whether or not a sane, rational person in 1780 meant that each individual should have guns or that the people should be able to raise a non-Federal military force because the specific meaning is right there in clear Greek.
As for actual laws, it should be a 100% bullet-proof legal defense that the law is invalid because it doesn't accomplish its scope. The scope says the law attempts to implement certain measures and for a certain purpose; if either the measure implemented in the part of the law you broke isn't in the scope statement *or* the action you took doesn't fit with the reasoning behind the law (for example: operating your cellphone GPS while parked at a light does NOT pose a danger to others through distracted driving increasing the risk of mis-operating a moving vehicle), the law is not applicable and has no standing.
Remember: The law must satisfy the scope as explained in Latin and Ancient Greek. Any conflict in the English scope statement with these is overruled by these. The Latin and Ancient Greek must not conflict with one another. The Latin and Ancient Greek may be partial clarifications and explanations of the English scope statement, explicitly yielding specific parts of the English scope statement to interpretation. They may even cite that a certain part of the English scope statement cannot be sufficiently explained in Latin or Greek, put bounds on the scope, offer guidance, and yield remaining interpretation as does not run afoul of these bounds and guidance. The full text of the law is in English, but is not valid outside the bounds of this scope.
Are we cool yet?
The economic cost of a minor traffic jam caused by a 5 second delay in moving your stopped vehicle and NOT getting a ticket because the police made a judgment call that you're not doing sufficient harm is lower than the economic cost of the PD processing a ticket, the driver having to go to court and take up the court's time, the driver having to expend several hours of their time in travel and time in court, records having to be expunged, etc.
It's also, amusingly, lower economic cost than fining the driver and/or pointing their license--which means, in the strictest sense, there is a net societal gain from not issuing citations like an autistic retard to drivers parked at a traffic signal using their phones!
Me and my friends actually devised a game where we go speeding along the highway at night (the only time speed cameras are active) and jumping red lights on open, clear intersections where you can verify it's safe. The cars have high-power dipole canons mounted and we try to EMP the cameras just as we get close. The more daring you feel, the closer you get before reducing the electronics to a smoldering heap.
Actually it does matter. I have had cars stall and fail when attempting to accelerate from a stop. Discovered the spark plugs got fouled because of a blocked EGR valve. The car ran fine until that moment.
It is 100% the responsibility of the driver who collided with the obstruction for colliding with a non-moving obstruction. The person sitting at a light could be cited for obstruction of traffic; however that does not move the blame for the collision onto him. A person sitting at a light being a dufus could easily be a disabled car, a cop car, a deer, a construction worker, a driver experiencing a heart attack, or any number of things that the next driver needs to remain aware of and react appropriately to.
Or are you the type that also blames gun manufacturers for murders?
No, he's not doing his job.
The job of a police officer is to ensure the smooth and peaceful running of society. Police officers are supposed to use their better judgment to decide if they need to intervene in an unsafe situation. They're not supposed to be walking porcine bureaucrats looking to randomly drain and damage society by mindlessly misapplying draconian rules into incidental situations.
Texting while driving laws were put into place because of the extreme danger of distracted driving. That danger isn't present when parked at a light--you might annoy someone by not moving when the light changes, and you'll obstruct traffic in a non-dangerous manner. We have accepted the danger of people referencing, but not programming, their GPS while driving; we certainly haven't targetted GPS use while parked at a light. Ticketing people for these things is inappropriate, regardless of what the law actually says. The law was put in place specifically to address certain societal problems; these actions do not intersect with those problems, and so the officer should apply his legal discretion rather than acting like a predatory dickhead.
PostgreSQL has settings that make up that last bit of reliability, putting it roughly on par with MongoDB.
And...
Short version: MongoDB allows you to, on a per-query basis, write data into the database at any level of reliability that MySQL and PostgreSQL provide. Single-server Journaled (WAL log shipping, WAL asynchronous streaming, MySQL master-slave replication), multi-server Replica Acknowledged (PostgreSQL WAL synchronous streaming), and a single-server "Acknowledged" mode that is faster but gives a weaker data durability guarantee (transaction is valid, not yet to disk, and not replicated).
"Extremely unlikely" comes out because the data's been replicated enough that it should be propagating further. Notably, 50%+1 nodes, so 2 nodes in a 3 node cluster, or 3 of 5, or 5 of 7. The data is also flushed to disk every 100mS ("j" write concern on a single node flushes the journal immediately and won't acknowledge until it's on disk). Once the data is to disk, it's 100% guaranteed to be there if the server experiences a recoverable system failure (power drop, kernel panic); it will be lost if it experiences filesystem corruption or a hard disk failure.
If you have more than 50% of the replica set go down, the whole cluster stops. The cluster is now refusing writes. If a replica node comes up with the data on disk, it will successfully propagate through the replica set and the data will not be lost.
To fail this, you need every replica node to acknowledge that it's got the data and then immediately fail in the 100mS window between acknowledging the query and writing it to disk; or you need them to suffer fatal hard disk failure or file system corruption and a system halt before they forward the data on to the others in the replica set. All of them. If you have a set of three, you need the primary and one of the secondaries to both fail; the primary won't acknowledge the transaction at all until 50%+1 of the replica set has it (one secondary in a replica set of 3), so the window of failure there is actually quite small,
As you can imagine, this involves a very carefully orchestrated exact sequence of catastrophic events each with low probability of occurring over long timescales. A system failure in any given 8.76 hour period has a 1 in 1,000 chance in a system that has 99.9% uptime. We're talking about two systems failing in exact windows of 0.1 seconds in tandem; two systems failing in the same 8.76 hour period has a probability p=1/1,000,000 with 99.9% uptime. There are 315,569.26 0.1 second periods in each 8.76 hour period, giving a probability of failure in any of these at the exact moment required of 1 in 315,569,260; probability of two exact failures in the correct way for a replica set of three is 1 in 99,583,957,900. With 99.99% uptime, it's 1 in 99,583,957,900,000,000. Note that the window is actually smaller than 0.1 seconds because it'll occur between a journal flush most likely, so it's more like a 50mS window on average and the numbers I gave should be doubled (1 in 2n).
It's a small window. PostgreSQL's synchronous WAL replication is 100% guaranteed if it acknowledges the transaction, because it doesn't acknowledge until both servers hit the disk; the only data loss possible is the even less likely case of both servers destroying their disks completely at the same time (actually this is probably more likely: the window of data loss would be, at a minimum, the time it takes to react to a failover by running an immediate automated backup; and even then, you're no longer streaming new data to another server). In a three-way replica set on PostgreSQL with synchronous WAL (Master and two Slaves), you need three such failures and thus your risk of data loss is slightly lower than MongoDB (again, because that third server will 100% guarantee data to disk before you get a successful COMMIT; but then you've got a wider window for disk failure on the last one).
Probability of failure is low.
No, I mean MongoDB will take a 3 database cluster and let you "Replica Acknowledge" a transaction with "Majority" count. Once it hits 50%+1 servers, it's 100% guaranteed solid unless you lose both servers. If both servers suffer a power drop at that point, the last server refuses to accept writes; when those servers come back, they will replay their oplog back to the last server to synchronize it. There's one flaw here: there's no "Replica Journal Acknowledge", so it's theoretically possible to lose that transaction anyway; both servers have to suffer a system failure (power drop, kernel panic) within 100mS of receiving the operation, since they write out their data to disk every 100mS. In practice this is extremely unlikely.
That means once you've sent it and gotten back that it's written, it is written. You'd have to lose both (or more--3 servers in a cluster of 5, etc.) servers' power or hard drives (corruption, failure) before the data is propagated further.
By contrast, Percona and MariaDB have XtraDB. XtraDB does optimistic locking: in normal autocommit, the transaction might get rolled back silently--it will write successfully to one server and return success, but if another server simultaneously gets a write that conflicts and starts propagating it then the transaction will be silently rolled back (i.e. undone, removed, lost, failed). With BEGIN-COMMIT transactions, you may get a Deadlock on "COMMIT" and then you're informed that it did in fact roll back the transaction and you must re-submit (i.e. do this if you actually care about durability of the data). With autocommit, as well as with any transactions (even explicit COMMIT) on MySQL master-slave replication or PostgreSQL WAL replication, you may in fact be informed that the transaction is 100% committed and then have that server FAIL and the slave comes up without that transaction--unavoidable silent data loss.
The failure mode expressed by MySQL master-slave replication and PostgreSQL WAL replication in the default asynchronous streaming replication mode is the same failure mode as with "Journaled" write concern in MongoDB. When running "Journaled" rather than "Replica Acknowledged," you write to exactly one server and are told it's committed when it's written to disk--it's durable on that server, but not necessarily replicated. If that server power drops and comes back up, it may find new operations have made its non-replicated operations invalid; it will then silently roll those back.
Therefor, in cluster layouts, it is possible for MongoDB to have a negligible reliability advantage over PostgreSQL's most common replication methods. PostgreSQL has settings that make up that last bit of reliability, putting it roughly on par with MongoDB. MongoDB has a guaranteed "It has reached enough servers that it is valid on the cluster unless God hates you" write concern by which the data is likely to actually be there if it tells you it's there, unless a very specific subset of servers experience a catastrophic failure in an extremely small (tenths of a second) window--a subset large enough to take down your entire cluster.
Short version: MongoDB allows you to, on a per-query basis, write data into the database at any level of reliability that MySQL and PostgreSQL provide. Single-server Journaled (WAL log shipping, WAL asynchronous streaming, MySQL master-slave replication), multi-server Replica Acknowledged (PostgreSQL WAL synchronous streaming), and a single-server "Acknowledged" mode that is faster but gives a weaker data durability guarantee (transaction is valid, not yet to disk, and not replicated).
*"PostgreSQL streaming replication is asynchronous by default. If the primary server crashes then some transactions that were committed may not have been replicated to the standby server, causing data loss. The amount of data loss is proportional to the replication delay at the time of failover."
MySQL is a piece of shit. MongoDB and PostgreSQL, depending on the data layout (MongoDB if you're looking for an indexed XML-YAML-JSON type of data with reliable data integrity in a cluster; PostgreSQL if you're looking for an indexed CSV with good replication and only minimal probability of basically silent data loss in a fail-over scenario).
Nigeria, although talking about Lagos is unfair since it's basically New York City, but in Africa.
These are not poor countries. "Poor countries with the greatest natural resources wind up with the most corrupt regimes" implies that the country started poor and gained an evil tyrant, not that it's poor because of the evil tyrant.
The whole statement is ridiculous doublethink. "Poor countries with the greatest natural resources"? In Burkina Faso, you can get gold by sifting tiny, tiny flecks out of dirt. That's not super-rich great natural resources; there's gold in the dirt and it takes a ridiculous amount of effort to get to it, so that's essentially "resource poor".
If these countries had great natural resources, they would be rich as living fuck. Don't tell me that ubiquitous presence of trace elements means "great natural resources", because that's like saying the bits of water you can squeeze from plants in the desert count as "well-hydrated marsh region." Hell, the desert's better off: you can squeeze a cactus. Imagine that water being distributed evenly across the desert as moisture in soil 6 inches under the sand. Water like a raging river, but you have to acquire it by concentrating what is an unending puff of dampness stretched across the vast and endless desert. More water than in the Mississippi, but at least we could stick a bucket into the Mississippi and get something vaguely drinkable.
They're poor for a reason. It's the exact opposite of having great natural resources.
Right. Some of these even have fish ladders because salmon. Hydro isn't zero-impact though.
Yeah my bad. 437GW was like a spec for the whole world or china or something. The biggest is an 8GW in Japan.
No, the issue is that a wind turbine produces a lot less than a single, large nuclear power plant. Nuke plants produce several gigawatts if they like; a single wind turbine produces 1-3MW, so if you have 2,500 wind turbines you can kind of approach what one fairly large nuke plant produces. Alta Wind Energy Center has 490 turbines and produces 1.3GW. Kashiwazaki-Kariwa produces 8GW. The US has 90 gigawatts of nuclear power (790TWh produced in 2011) supplied by about 100 small and large nuclear plants.
People are dumb enough to think nuclear power is dangerous. There was a huge news story about a nuclear plant leaking radiation into the water table a few years ago. You know how much radiation it leaked? Enough over 10 years that if you extrapolated the leak by 10,000 years and concentrated that into one acute dose, you might have some long-term health effects a few decades down the line.
Radiation is a big scary boogieman. It's like fire. Oh my god there's a fire! Well if you stick your face into the charcoal grill, you might have some trouble; but overall it's .. warm, a few feet away... and if you're any appreciable distance off it's not so warm. That warmth won't hurt you. If the thing tips over and sets fire to your house and you're entrapped in the blaze, that's different.
People imagine that "it's warm downwind because a concentrated column of air is blowing my way" is the same as "HELP I'M ON FIRE!!!" They get a little radiation leak, or even a failed plant, and they're like... oh god, contamination, contamination everywhere! But the truth is nothing changes, the minimal amount of radiation exposure is largely untroubling, and everything is fine. Unless it's Tjernobyl spewing liquefied toxic waste everywhere and dumping raw liquid nuclear fuel into the water table and scattering tons and tons and tons of radioactive matter into the air, it's a non-starter.
I mean for god's sake, people in Taiwan living in highly radioactive buildings (1000 times background radiation) show 40% lower incidences of cancer and overall have better health; most studies only manage to conclude that we can't decide if the reduced incidence of ALL TYPES OF CANCER (not Simpson's Paradox; all means each individual type--less leukemia, less lung cancer, less skin cancer, etc.) is because the population is younger or because these are mostly rich upper-middle-class people who lead healthier lifestyles. That's a choice between "no real change" and "These people don't eat garbage and have heard about something called 'jogging'".
A radiation plant with a leak will not expose the local population to 1000 times background radiation. It won't expose them to a short-term 1000x dose or a lifetime of 1000x background radiation. Even a major nuclear disaster, unless the plant fucking explodes, probably won't expose the local population to doses like that--what do you think containment buildings are for? In most cases, a leak that raises alarms and exceeds regulatory guidelines and gets the NRA on their asses will expose the population 100 feet away from the leak to less background radiation than living within 10 miles of a coal plant does. And even then, sometimes it's fucking harmless (all this bullshit over a minor TRITIUM leak?!).
Big bogeyman. Nuclear plants aren't as scary as people want you to believe. Look at the Fukushima disaster--people are talking about the disaster, but not about the actual impact. We've arbitrarily assigned a high impact to this disaster, but is there any real impact at all? Sure. I bet some people get cancer earlier than they would normally. Eventually. We'll see in 30 years.
Uh, a thousand times more people would be employed by the wind farm. How do you stand up a 437GW wind farm?
Minimal environmental impact hahahahahahahaha....
That's rich. Flooding an area has "minimal environmental impact". It only ever really worked at Niagra Falls, since you didn't need to create a giant fucking reservoir where there was a wilderness.
I bet the people of Fukushima suffer minimal hardship. Like, a third of them get cancer, when they're fucking old, and probably would have cancer anyway.
Argumentum ad absurdum is useful when you're trying to make a point. In some cases it is patently absurd to "stick to principles".
In some places in Africa they have effective slave labor. People are "free", but children as young as 5 or 6 work in mines with no safety equipment laboring to dig up little flecks of gold. They develop lifelong health issues. Many die. They do this because resources are so scarce that your whole family right down to your young children must work in such conditions to make enough money to spend every day hungry, but alive with just enough food to keep you sickly and breathing. This isn't just a small segment of the population; entire regions operate like this, with just about everyone scraping by. Tanzania, Burkina Faso, places like that. Merchants show up where the mines are bearing nothing beyond cheap food and tents to sell it from, because these people can't go very far to eat.
Capitalism. Capitalism in a region too poor to support itself.
Capitalism works in America, in Canada (OIL RICH!), in Norway (OIL RICH!), in Japan, in the UK, in France (France has an interesting system that I like, by the way, and it works extremely well for them)... all places where you can sleep in, go to work, and then stop at McDonalds and get a really shitty hamburger.
Capitalism isn't working for some significantly sized regions in Tanzania. Capitalism isn't working for Burkina Faso. And somehow these people can't just leave and go to the adjacent region where they're truly free, and not slaves of necessity. Would that it were so simple to walk five miles to freedom; but apparently five miles is either cripplingly far or there is nothing five miles away any better than what's directly here.
These people would be, to a feudal lord, important human assets. These serfs are valuable. They are not valuable to gold miners because we can always find more; but as an asset that comes with the land, they are valuable to a landowner. When your serfs are sickly, they don't work as well; when they die off, they expend time and energy on dying and death rites instead of working. You must improve these conditions, must find other ways for them to produce, must produce medicine to keep them healthy and technology to make them harvest the resources on your land more effectively to make you richer.
Gold mines are interesting. You have to mine out an area and move to the next before the next guy, so mining at speed is important. On the other hand, larger operations that don't have access to essentially slave labor can't do that (i.e. in the US or Europe); we've developed ways to mine gold by using plants, which takes about 10 years. Now if the landlord owns the land, his only possible way to enrich himself is to acquire more land; he doesn't care if it takes 10 years to mine an area by phytomining (including by digging and then phytomining the pile of dirt, rather than spending a lot of time sifting), because he owns the land he owns. If the serfs can spend their time growing food on non-mine land and building exportable things, living healthier, they're more productive and thus making the landlord richer. Automated mechanical sifters are attractive too.
So take your pick: small children choking on mineral dust, barely surviving to breed, taken as fodder by opportunists in these resource-poor regions; or feudalism, serfs, some semblance of value in the people, and a lot of hardship because resource-poor has a real meaning and we can't just give you a magical solution like "Captialism" or "Feudalism" or "Marxism" and create an instant utopia.
By the by, an interesting thought about Burkina Faso: The whole of Europe and the New World are trying to boycott any gold from Burkina Faso, desperately trying to trace gold sources back to avoid anything from there. They say the conditions there and the child slave labor are intolerable, so they're just going to break the backs of the slave drivers. Once they do that... ...there
It's cooler than when I read it as "Find millions of nearby assholes" and decided that Craigslist and OKCupid already do this. Or living in Boston or Baltimore, where you just have to look outside.
An enslaved population must survive to have a revolution to gain its freedom back; a free, poor society must survive to develop into a free, prosperous one. Principles are easy to fall back on: just refuse to negotiate with terrorists while they continue walking down the line of hostage 7 year old girls and mouth and anal raping each of them in turn trying to break you.
Non-negotiation with terrorists has real impact. Sometimes people die who didn't need to; but the ineffectiveness of hostage negotiation means fewer people are so inclined. Those that are will raise the stakes, which means that the president's daughter will be kidnapped by ninjas instead of just some random schoolgirl. Although a few suffer today, the many will benefit from their suffering and death: fewer will ever be taken hostage, and thus fewer will be killed as hostages in the future. This is a good place to stand on principle.
Until they're going to nuke seven elementary schools at once.
In a society so poor that "freedom" and "capitalism" results in the extreme suffering and death of 98% of the population--thousands, millions--some control of resources is better. A feudal society is only quasi-slavery: you're effectively "free" in that you can live your life as long as you produce--which is like America, where you're only "free" until you're a lazy hippie who doesn't get a job and tries to smoke pot all day and leech the system that's there for people who are TEMPORARILY down, not permanent willful parasites. In feudalism, obviously you can't go on vacation; you can't leave your land. On the other hand, extremely scarce resources are managed somewhat better via feudalism, so folks get by under appreciable hardship rather than starving and rotting as a society. Sometimes the hardship even gives way to a tolerable, somewhat enjoyable, but not very exciting way of life.
Freedom comes at a price. If you can't afford it, then we take death in exchange.