...but that's exactly what the ruling does. The original case was a businessman objecting to Google links to newpaper stories about his life. This is no different.
Fact is, the court that issued this ruling screwed up big time. Perhaps, if Google can find a few more egregious deletions to make, the European Parliament will correct the error.
I think the big problem here is that Google are expected to be the judge, jury and executioner and are getting smacked down when someone thinks they made the wrong judgement call. This stuff should be going to an independent judge instead of expecting Google to uphold a new law that has a fairly vague scope.
How do they choose the exchange? Government property must be auctioned off to the highest bidder, otherwise they are favoring a business over others.
One of those little things that they do to maintain the appearance that they are not corrupt.
I'm curious how they handle foreign currency which is seized - if they seize a truck full of euros, do they auction them or of just exchange them for dollars? If the latter, what's the difference between that and doing the same with bitcoins?
Home owners can't really lose with solar PV, unless they somehow get screwed on workmanship or installation costs. The panels with always pay for themselves in a few years and it's shear madness that new houses are being built without it.
If you're going to live in the house for at least the break-even time then yes, you probably can't lose. However, I'm less convinced that it adds so much to the value of the house: if there are 2 identical houses for sale, but one of them has a brand new £20K installation of PV panels on the roof, are people really going to pay £20K more for that one? I suspect not, because its an up-front cost and some people simply won't be able to afford that much up-front. (Ok, so people will tack it onto the mortgage, but that means convincing the bank to give you a bigger mortgage).
So that is why new houses aren't built with PV panels - because it almost certainly doesn't raise the sale value of the house by the amount spent on the panels.
In general, solar panels work well for the rich but not so well for the poor: If you've bought the house you're going to live in for a significant number of years and you can afford the large up-front cost of the panels then it's a good investment. But only the richer part of the population can do this, so subsidising solar power actually just ends up transferring wealth from the poor to the rich, which is why it's contraversial.
Data mining and intercepting bad company experiences and "making good" on them. For example, we had something delivered via UPS. The driver left it on our front step, didn't ring the doorbell, and just left. It sat out there for hours before we realized it was there. The package could have easily been stolen during that time and neither UPS nor I would have known until it was much too late. We complained on Twitter and UPS contacted us in an attempt to find out what went wrong and how they could improve their policies.
I think this is a bit of a red herring. If the company cares about customer service then you shouldn't need to complain on a public forum to get them to pay attention - they should provide a customer support line and actually take all calls to it seriously.
Spot bad experiences, help minimize bad PR by helping those customers, and minimize future bad PR by fixing those problems before more customers are affected.
Now you've hit the nail on the head. Social media is actually a problem for companies: prior to social media, a company screws up, you complain, they ignore you, you moan about it to about 3 of your mates. Now, a company screws up, you complain, they ignore you, you moan about it on twitter, it goes viral and a hundred thousand people see how badly the company treated you. The company isn't interested in helping their customer (if they were, they would've helped their customer whether or not it turned into a PR disaster), instead they are interested in avoiding really bad PR - really bad PR that wouldn't ever have happened without social media.
I don't use Twitter, but I have sometimes wondered whether its worth getting an account just to make complaints, since complaints on Twitter seem to attract a much quicker and more helpful response than calls directly to customer services.
Picking up an unnatural, focused emission of energy from another planet would already be pretty good, even if we cannot decode the signal or even recognize it as such.
How do you determine what is unnatural? Over the years astronomers have picked up *lots* of signals that had no natural explanation at the time but do now...
Personally I think the whole thing is likely a waste of time - we've only been using radio for interplanetary communications for a few decades and things are now rapidly moving towards laser communications. Assuming another civilisation follows a similar path, the time between "not advanced enough to detect" and "too advanced to detect" seems pretty short.
The absurd release frequency, the unnecessary changes, and the bad quality forced me to air-gap my system and freeze it in an ancient version in order to keep it running (or, better said, in order to reduce the risk of it breaking down). I stopped recommending fedora ages ago. Now that that system fulfilled its original purpose, it will be repurposed and updated with something different, probably CentOS or Mint.
I think you're rather missing the point of Fedora. The whole point is a Free, rapid release cycle distribution to track the (b)leading edge technologies. The good stuff that drops out of this goes into RHEL a few years later, whilst the bad stuff is abandoned. If you wanted a long-term-support distro, why did you choose a rapid release cycle one in the first place? RHEL, CentOS or Scientific Linux are much more sensible if your're not interested in the latest features; but you can't have both - you can't have the latest stuff that was only developed last month unless you go with a rapid release cycle distro.
Are there any benefits that a random British person could point out, that are the result of UK being in the EU?
Economic advantages, of course, but also a whole swathe of good laws have come from the EU. The anti-EU crowd always like to point at the bad laws (and of course, there are some) as a reason to leave whilst completely ignoring all the good laws that are only here as a result of the EU.
I don't see how a conviction for possessing child porn is irrelevant or outdated. So I don't like his chances.
If the sentence has been served then is it really in anyone's interest to keep persecuting someone for an crime that they once committed?
If someone is still a danger to the public, they shouldn't be allowed out in public unsupervised. If they aren't a danger to the public then the public doesn't need to know.
For the same reason your email server accepts emails with fake sender addresses - it's usually not possible for the telco to know that its fake.
Analogy fail. Emails are not billed as such, they're just part of the sea of data flowing across the network. Phone calls, on the other hand, *are* discretely billed, so phone companies *must* have an accurate record of where calls are coming from so they know where to send the bill.
Telcos know where the call entered their network. On the originating network, this means that they can accurately know which subscriber line it came from and bill as appropriate. On transit / destination networks, the identity of the originator is not known with confidence, nor does it need to be for billing purposes (the destination network is not billing the original subscriber; they may be billing the network that passed the call on to them (originator or transit), and of course they know which network that was, but not the identity of the actual originating subscriber.
It's called ANI (automatic number identification). It's not Caller ID and is not normally spoofable.
AFAIK there is no requirement for a network to pass ANI data to another network when a call crosses between them. It certainly isn't required for the operation of the network itself (the signalling traffic is routed by point code and the media is identified by circuit - the only telephone number actually required for call routing is the callee's and even that is only required for call setup; of course, SS7 networks are not transparent end-to-end networks like the internet, so the "originating point code" that the callee sees isn't going to be the actual originating point code, especially if the call originated off-network).
Heck, VoIP gateways aren't usually going to provide a meaningful ANI anyway.
Mine doesn't. Its fairly easy to setup your mail server to only accept mail from properly configured mail servers, in which case you can ensure the message came from a server that should be responsible for sending you a message from that address.
Most of the time there is absolutely no way to know which mail server is responsible for sending mail for a particular address. People can publish SPF records which provide you with this information, but very few people do, so you can't rely on this.
You can look at the MX records to see which servers can receive mail for that domain, but that doesn't tell you which can send mail, so again you can't rely on this at all.
You can do sender verification callouts, but this only confirms that the address is valid, not that the message you are receiving actually came from it. Also, sender verification callouts are considered a Bad Thing, since they can be abused to create a reflection DDoS attack.
If you want to set up a mail server that will only accept mail from systems that publish SPF records then by all means you can do so and you'll massively cut the amount of spam you receive. You'll also massively cut the amount of legitimate mail you get too.
Why are the telcos accepting fake caller IDs, and why are the law enforcement agencies tolerating this crap from the phone companies? That's the problem. The police are, as you say, just reacting to the call they get. The question is why the caller is allowed to lie about what number they're calling from.
For the same reason your email server accepts emails with fake sender addresses - it's usually not possible for the telco to know that its fake.
Well, let's cover the office and the shopping center and the parking garage with solar panels. At least some of it could come from the sky at the point of use. And if you're going to run a lot of capacity there anyway so that cars which are there can be charged, it's a good place to site the panels even when they're not being used locally.
Offices and shopping centres already use a lot of power. I'm not sure why the existence of electric cars really changes anything - they haven't installed PV to cover their usual energy consumption so why would they install PV to charge your car?
Besides - if the camera takes 0.1W to record then it takes 0.1W - all reducing the footage quality does is reduce the amount of RAM needed as a buffer.
Not at all. The amount of power used by CMOS hardware is basically proportional to the number of transistors that are being switched, and how frequently they are switching. So each time you capture a frame you have to:
- reset the sensor's pixels
- read the sensor's pixels
- amplify the signal
- debayer the data
- possibly compress the data
- store the data somewhere Each of these steps will take a certain amount of energy. Obviously the more frequently you capture a picture, the more frequently you have to do all of the above and so the power consumption increases. There's a reason why your laptop or phone puts the CPU to sleep between operations, and it's the same reason why your computer gets hot when asked to do more work.
I was confused about how someone could be charged for access to "open source" information..
Open source and public domain are not the same things - most open source data is copyrighted and made available through a suitably permissive licence. Break that licence and you can be sued just as easily as if you were breaking a closed source licence.
I keep having conversations with my students where I explain why they shouldn't pirate books, or at least should make sure that the authors are getting paid (for instance, buying a legal copy then pirating / cracking it if it has DRM to get a useful one.)...and yet I have a lot of trouble trying to work up enthusiasm for telling them not to pirate textbooks.* Particularly problematic, as I've shown a few how to torrent. (Heck, I've shown faculty members how to torrent.)
* As opposed to professional reference books.
I'm not a big fan of copyright infringement, but I do wonder why you would pay good money for something you're going to have to break the law to use rather than just breaking the law to use it without paying first...
In an accident in which a fatality is unavoidable, I don't want my car to agree with the other cars that its best for me to be the fatality
That's shortsighted, you're assuming you will always be on the losing side of this situation. If the cars strategy is to preserve as many lives as possible, even if that means harm to the occupant, you are more likely to be saved by this strategy when it is the other drivers car trying to save you.
My purchasing decision makes no difference to other peoples' purchasing decisions - if I decide to be selfless and buy a self sacrificing car and everyone else buys a self preserving car then the only people better off are the people who aren't me. I would want the things I can make decisions about to improve my chances of survival - I can't control what other people's decisions are (but I think its reasonable to assume that, given the choice, other people would also choose a self-preserving car too). This is basic game theory stuff - this is every man for himself, the only way you're going to get everyone to make the selfless decisions is if you force them to do so.
There's no such thing as an intentional accidents. An autonomous program that is paying attention will not have such a situation and therefore the manufacturers will always be responsible for failure.
If a car shoots out from a blind junction at speed and you can't stop in time, that's an unavoidable accident - the car could not be seen in advance, so the autonomous program couldn't have avoided the accident even if its paying attention the whole time. You could argue that you should be going slow enough that your stopping distance is short enough to avoid the collision, but on a lot of roads this would seriously hinder traffic flow - at some point you just have to trust that other drivers are following the rules of the road and accept that the risk can't be completely eliminated.
Similarly, mechanical failures can't always be predicted - you're overtaking someone and their wheel comes off causing them to swerve into you. Impossible to predict so now you're left trying to reduce the seriousness of the inevitable accident. Hell, your own car may have a mechanical failure that the computer couldn't detect.
Braking power isn't infinite. Wheel braking will eventually skid the wheels (which is why we have anti-lock brakes now, so you can still steer while braking). Are you thinking cars should be equipped with dragster-style parachutes, or retro-rockets? Or just a bloody great anchor that the computer can deploy and tear up the road?
Even when the car has deployed the parachute, the anchor, and the retro-rocket is still firing, the computer might still not be able to stop going into that tree that's just fallen over. Plus all those negative G forces are going to smear the drivers eyeballs over the inside of the windscreen.
The g forces of retro rockets is probably going to smear the driver over the inside of the windscreen rather less than hitting the tree at full speed though:)
Options would have to be costed. Many things would feed into that. The problem of course is that for all of those costings, probability multiplied by survivability does not produce a linear outcome of quality of life value; you could assign a value of harm to each individual present, but you could not get a meaningful figure by summation.
Don't forget that the computer must make sure that the bit of the car that gets crumpled is the highest profit margin component for the dealer to replace:)
There's an interesting idea - if you networked each of the cars and they shared a common utility function (i.e. the thing that determines how "good" or "bad" each possible result was) they could reach a common consensus on what the "globally best" course of action was.
Networking cars seems like an excellent idea until you consider the security problems - its pretty hard to design a system that can pass enough information around to be worthwhile, without also allowing untrustworthy systems to pass misinformation.
Also, I think its unrealistic to expect cars to agree on a "globally best" course of action, since the public isn't going to want to have an autonomous car that doesn't have some sense of self preservation. In an accident in which a fatality is unavoidable, I don't want my car to agree with the other cars that its best for me to be the fatality.
I'm reminded of Michael Sandel's televised series on ethics.
If you could stop a runaway train from going over a ravene, by pulling a lever, thus saving 300 people, but the lever sent the train down a different track on which 3 children were playing, what do you do?
Somehow, involving innocents seems to change the ethical choices. You're no longer just saving the most lives, but actively choosing to kill innocent bystanders.
Well, part of the point of that thought experiment is to demonstrate that people usually have more problem with bad stuff happening through their actions than through their inactions.
Slashdot Mirror
...but that's exactly what the ruling does. The original case was a businessman objecting to Google links to newpaper stories about his life. This is no different.
Fact is, the court that issued this ruling screwed up big time. Perhaps, if Google can find a few more egregious deletions to make, the European Parliament will correct the error.
I think the big problem here is that Google are expected to be the judge, jury and executioner and are getting smacked down when someone thinks they made the wrong judgement call. This stuff should be going to an independent judge instead of expecting Google to uphold a new law that has a fairly vague scope.
How do they choose the exchange? Government property must be auctioned off to the highest bidder, otherwise they are favoring a business over others.
One of those little things that they do to maintain the appearance that they are not corrupt.
I'm curious how they handle foreign currency which is seized - if they seize a truck full of euros, do they auction them or of just exchange them for dollars? If the latter, what's the difference between that and doing the same with bitcoins?
Home owners can't really lose with solar PV, unless they somehow get screwed on workmanship or installation costs. The panels with always pay for themselves in a few years and it's shear madness that new houses are being built without it.
If you're going to live in the house for at least the break-even time then yes, you probably can't lose. However, I'm less convinced that it adds so much to the value of the house: if there are 2 identical houses for sale, but one of them has a brand new £20K installation of PV panels on the roof, are people really going to pay £20K more for that one? I suspect not, because its an up-front cost and some people simply won't be able to afford that much up-front. (Ok, so people will tack it onto the mortgage, but that means convincing the bank to give you a bigger mortgage).
So that is why new houses aren't built with PV panels - because it almost certainly doesn't raise the sale value of the house by the amount spent on the panels.
In general, solar panels work well for the rich but not so well for the poor: If you've bought the house you're going to live in for a significant number of years and you can afford the large up-front cost of the panels then it's a good investment. But only the richer part of the population can do this, so subsidising solar power actually just ends up transferring wealth from the poor to the rich, which is why it's contraversial.
Data mining and intercepting bad company experiences and "making good" on them. For example, we had something delivered via UPS. The driver left it on our front step, didn't ring the doorbell, and just left. It sat out there for hours before we realized it was there. The package could have easily been stolen during that time and neither UPS nor I would have known until it was much too late. We complained on Twitter and UPS contacted us in an attempt to find out what went wrong and how they could improve their policies.
I think this is a bit of a red herring. If the company cares about customer service then you shouldn't need to complain on a public forum to get them to pay attention - they should provide a customer support line and actually take all calls to it seriously.
Spot bad experiences, help minimize bad PR by helping those customers, and minimize future bad PR by fixing those problems before more customers are affected.
Now you've hit the nail on the head. Social media is actually a problem for companies: prior to social media, a company screws up, you complain, they ignore you, you moan about it to about 3 of your mates. Now, a company screws up, you complain, they ignore you, you moan about it on twitter, it goes viral and a hundred thousand people see how badly the company treated you. The company isn't interested in helping their customer (if they were, they would've helped their customer whether or not it turned into a PR disaster), instead they are interested in avoiding really bad PR - really bad PR that wouldn't ever have happened without social media.
I don't use Twitter, but I have sometimes wondered whether its worth getting an account just to make complaints, since complaints on Twitter seem to attract a much quicker and more helpful response than calls directly to customer services.
Short of a bloody revolt, what exactly can the citizens do about it?
They can stop voting for people who pander to their fears.
Who exactly do you suggest? Coz the only way I can see to do as you say is to simply not vote for anyone, and I can't see how that's going to help.
And the global dictatorship is slowly being pieced together.
And citizens do nothing, amazingly. People with any knowledge of history should be scared shitless - I know I am.
And soon it'll be too late to do anything about it...
Short of a bloody revolt, what exactly can the citizens do about it?
Picking up an unnatural, focused emission of energy from another planet would already be pretty good, even if we cannot decode the signal or even recognize it as such.
How do you determine what is unnatural? Over the years astronomers have picked up *lots* of signals that had no natural explanation at the time but do now...
Personally I think the whole thing is likely a waste of time - we've only been using radio for interplanetary communications for a few decades and things are now rapidly moving towards laser communications. Assuming another civilisation follows a similar path, the time between "not advanced enough to detect" and "too advanced to detect" seems pretty short.
The absurd release frequency, the unnecessary changes, and the bad quality forced me to air-gap my system and freeze it in an ancient version in order to keep it running (or, better said, in order to reduce the risk of it breaking down). I stopped recommending fedora ages ago. Now that that system fulfilled its original purpose, it will be repurposed and updated with something different, probably CentOS or Mint.
I think you're rather missing the point of Fedora. The whole point is a Free, rapid release cycle distribution to track the (b)leading edge technologies. The good stuff that drops out of this goes into RHEL a few years later, whilst the bad stuff is abandoned. If you wanted a long-term-support distro, why did you choose a rapid release cycle one in the first place? RHEL, CentOS or Scientific Linux are much more sensible if your're not interested in the latest features; but you can't have both - you can't have the latest stuff that was only developed last month unless you go with a rapid release cycle distro.
Are there any benefits that a random British person could point out, that are the result of UK being in the EU?
Economic advantages, of course, but also a whole swathe of good laws have come from the EU. The anti-EU crowd always like to point at the bad laws (and of course, there are some) as a reason to leave whilst completely ignoring all the good laws that are only here as a result of the EU.
I don't see how a conviction for possessing child porn is irrelevant or outdated. So I don't like his chances.
If the sentence has been served then is it really in anyone's interest to keep persecuting someone for an crime that they once committed?
If someone is still a danger to the public, they shouldn't be allowed out in public unsupervised. If they aren't a danger to the public then the public doesn't need to know.
Analogy fail. Emails are not billed as such, they're just part of the sea of data flowing across the network. Phone calls, on the other hand, *are* discretely billed, so phone companies *must* have an accurate record of where calls are coming from so they know where to send the bill.
Telcos know where the call entered their network. On the originating network, this means that they can accurately know which subscriber line it came from and bill as appropriate. On transit / destination networks, the identity of the originator is not known with confidence, nor does it need to be for billing purposes (the destination network is not billing the original subscriber; they may be billing the network that passed the call on to them (originator or transit), and of course they know which network that was, but not the identity of the actual originating subscriber.
It's called ANI (automatic number identification). It's not Caller ID and is not normally spoofable.
AFAIK there is no requirement for a network to pass ANI data to another network when a call crosses between them. It certainly isn't required for the operation of the network itself (the signalling traffic is routed by point code and the media is identified by circuit - the only telephone number actually required for call routing is the callee's and even that is only required for call setup; of course, SS7 networks are not transparent end-to-end networks like the internet, so the "originating point code" that the callee sees isn't going to be the actual originating point code, especially if the call originated off-network).
Heck, VoIP gateways aren't usually going to provide a meaningful ANI anyway.
Mine doesn't. Its fairly easy to setup your mail server to only accept mail from properly configured mail servers, in which case you can ensure the message came from a server that should be responsible for sending you a message from that address.
Most of the time there is absolutely no way to know which mail server is responsible for sending mail for a particular address. People can publish SPF records which provide you with this information, but very few people do, so you can't rely on this.
You can look at the MX records to see which servers can receive mail for that domain, but that doesn't tell you which can send mail, so again you can't rely on this at all.
You can do sender verification callouts, but this only confirms that the address is valid, not that the message you are receiving actually came from it. Also, sender verification callouts are considered a Bad Thing, since they can be abused to create a reflection DDoS attack.
If you want to set up a mail server that will only accept mail from systems that publish SPF records then by all means you can do so and you'll massively cut the amount of spam you receive. You'll also massively cut the amount of legitimate mail you get too.
Why are the telcos accepting fake caller IDs, and why are the law enforcement agencies tolerating this crap from the phone companies? That's the problem. The police are, as you say, just reacting to the call they get. The question is why the caller is allowed to lie about what number they're calling from.
For the same reason your email server accepts emails with fake sender addresses - it's usually not possible for the telco to know that its fake.
Well, let's cover the office and the shopping center and the parking garage with solar panels. At least some of it could come from the sky at the point of use. And if you're going to run a lot of capacity there anyway so that cars which are there can be charged, it's a good place to site the panels even when they're not being used locally.
Offices and shopping centres already use a lot of power. I'm not sure why the existence of electric cars really changes anything - they haven't installed PV to cover their usual energy consumption so why would they install PV to charge your car?
Besides - if the camera takes 0.1W to record then it takes 0.1W - all reducing the footage quality does is reduce the amount of RAM needed as a buffer.
Not at all. The amount of power used by CMOS hardware is basically proportional to the number of transistors that are being switched, and how frequently they are switching. So each time you capture a frame you have to:
- reset the sensor's pixels
- read the sensor's pixels
- amplify the signal
- debayer the data
- possibly compress the data
- store the data somewhere
Each of these steps will take a certain amount of energy. Obviously the more frequently you capture a picture, the more frequently you have to do all of the above and so the power consumption increases. There's a reason why your laptop or phone puts the CPU to sleep between operations, and it's the same reason why your computer gets hot when asked to do more work.
I was confused about how someone could be charged for access to "open source" information..
Open source and public domain are not the same things - most open source data is copyrighted and made available through a suitably permissive licence. Break that licence and you can be sued just as easily as if you were breaking a closed source licence.
Power requirements go down a LOT if you're writing to RAM instead of flash memory and not displaying anything on a video screen.
eg. I've seen CMOS sensors that use less than 0.1W.
It would also seem reasonable for the 30 second prebuffer to run at a reduced frame rate to save battery.
Ethics. He ants the author t get something.
I only assume he has never stepped foot into a library.
If the author wants to get paid, they can surely produce a product that can sensibly be used without breaking the law...
I keep having conversations with my students where I explain why they shouldn't pirate books, or at least should make sure that the authors are getting paid (for instance, buying a legal copy then pirating / cracking it if it has DRM to get a useful one.) ...and yet I have a lot of trouble trying to work up enthusiasm for telling them not to pirate textbooks.* Particularly problematic, as I've shown a few how to torrent. (Heck, I've shown faculty members how to torrent.)
* As opposed to professional reference books.
I'm not a big fan of copyright infringement, but I do wonder why you would pay good money for something you're going to have to break the law to use rather than just breaking the law to use it without paying first...
In an accident in which a fatality is unavoidable, I don't want my car to agree with the other cars that its best for me to be the fatality
That's shortsighted, you're assuming you will always be on the losing side of this situation. If the cars strategy is to preserve as many lives as possible, even if that means harm to the occupant, you are more likely to be saved by this strategy when it is the other drivers car trying to save you.
My purchasing decision makes no difference to other peoples' purchasing decisions - if I decide to be selfless and buy a self sacrificing car and everyone else buys a self preserving car then the only people better off are the people who aren't me. I would want the things I can make decisions about to improve my chances of survival - I can't control what other people's decisions are (but I think its reasonable to assume that, given the choice, other people would also choose a self-preserving car too). This is basic game theory stuff - this is every man for himself, the only way you're going to get everyone to make the selfless decisions is if you force them to do so.
There's no such thing as an intentional accidents. An autonomous program that is paying attention will not have such a situation and therefore the manufacturers will always be responsible for failure.
If a car shoots out from a blind junction at speed and you can't stop in time, that's an unavoidable accident - the car could not be seen in advance, so the autonomous program couldn't have avoided the accident even if its paying attention the whole time. You could argue that you should be going slow enough that your stopping distance is short enough to avoid the collision, but on a lot of roads this would seriously hinder traffic flow - at some point you just have to trust that other drivers are following the rules of the road and accept that the risk can't be completely eliminated.
Similarly, mechanical failures can't always be predicted - you're overtaking someone and their wheel comes off causing them to swerve into you. Impossible to predict so now you're left trying to reduce the seriousness of the inevitable accident. Hell, your own car may have a mechanical failure that the computer couldn't detect.
Braking power isn't infinite. Wheel braking will eventually skid the wheels (which is why we have anti-lock brakes now, so you can still steer while braking). Are you thinking cars should be equipped with dragster-style parachutes, or retro-rockets? Or just a bloody great anchor that the computer can deploy and tear up the road?
Even when the car has deployed the parachute, the anchor, and the retro-rocket is still firing, the computer might still not be able to stop going into that tree that's just fallen over. Plus all those negative G forces are going to smear the drivers eyeballs over the inside of the windscreen.
The g forces of retro rockets is probably going to smear the driver over the inside of the windscreen rather less than hitting the tree at full speed though :)
Options would have to be costed. Many things would feed into that. The problem of course is that for all of those costings, probability multiplied by survivability does not produce a linear outcome of quality of life value; you could assign a value of harm to each individual present, but you could not get a meaningful figure by summation.
Don't forget that the computer must make sure that the bit of the car that gets crumpled is the highest profit margin component for the dealer to replace :)
There's an interesting idea - if you networked each of the cars and they shared a common utility function (i.e. the thing that determines how "good" or "bad" each possible result was) they could reach a common consensus on what the "globally best" course of action was.
Networking cars seems like an excellent idea until you consider the security problems - its pretty hard to design a system that can pass enough information around to be worthwhile, without also allowing untrustworthy systems to pass misinformation.
Also, I think its unrealistic to expect cars to agree on a "globally best" course of action, since the public isn't going to want to have an autonomous car that doesn't have some sense of self preservation. In an accident in which a fatality is unavoidable, I don't want my car to agree with the other cars that its best for me to be the fatality.
I'm reminded of Michael Sandel's televised series on ethics.
If you could stop a runaway train from going over a ravene, by pulling a lever, thus saving 300 people, but the lever sent the train down a different track on which 3 children were playing, what do you do?
Somehow, involving innocents seems to change the ethical choices. You're no longer just saving the most lives, but actively choosing to kill innocent bystanders.
Well, part of the point of that thought experiment is to demonstrate that people usually have more problem with bad stuff happening through their actions than through their inactions.