The FDA does regulate "adverse event reporting." Look at any bottle of drugs you have (in the USA). Somewhere on it is a phone number. If you call that number and report that something bad happened after you took that pill you'll kick off a complex processes designed to ensure that all the details of your problem are captured, tracked, and ultimately reported back to the FDA. The FDA can require companies to study any trends that arise, and drugs actually have been taken off the market as a result.
The concern is that if you post "this drug made me sick" on a facebook wall, you just informed the company about an adverse event that the company is now legally required to follow-up on. However, to do this they might have to subpoena your personal information from Facebook, your ISP, and who-knows how many other intermediaries if you tried to make the account anonymous. They then need to try to track you and your doctor down and find out what happened. This is a legal nightmare. If you just called the number on your prescription bottle they'd gather the legally-required info up-front, and would know how to track you down.
I'd be a fan of just exempting social media from these kinds of requirements - if you want to moan about getting a stomachache on facebook have fun, but you shouldn't expect it to result in an investigation. However, government agencies aren't known for common sense, and they are known for beating people up over subtle interpretations of the rules.
However, I'd think that any tower hardware capable of serving up 3G data on both the T-mobile and ATT frequencies/protocols could probably handle CDMA/etc just as easily.
And you'd be wrong. At least not without spending assloads of money.
Can you supply details? Does ATT or T-Mo have lots of hardware deployed which could be re-tasked to serve the other networks 3G phones but not a CDMA-based network? My understanding is that the ATT and T-Mo 3G networks are fairly different (different frequencies and protocols), so the only way I figure the same hardware could support both is if it uses software-defined radio and has very wide multiband support, and hardware with these capabilities probably could handle CDMA. Then again, I could be wrong on that. And, I wouldn't be surprised if none of the towers out there could support both without hardware changes.
So, my employer does this kind of nonsense, and they have literally billions of dollars in "cash" on the balance sheet. So, the only cash flow issue is making the cash flow look good - not that they can't write checks.
And the MBAs have learned a long time ago that the goal isn't to make the company do well - the goal is to make your paycheck as large as possible. That does not require making the company do well - it merely requires convincing the boss that you're making the company do well, usually by reducing the size of the budget under your immediate control.
Yup, but the problem with your argument is that the kinds of savings you're arguing for are hard to measure. Sure, you can measure PC boot time, but what is the actual production output of a department after you fix their boot times? Chances are it would be hard to trace your improvement to real-world impact. I'm sure it is there and it pays for itself, but if you can't show it in numbers, then you can't get a bonus for implementing it. You can, however, get in trouble if you mess things up when implementing it. So, to an IT manager what you propose is all downside.
The whole IT-is-a-cost-center MBA argument would be recognized as being silly if applied to any other less-complex technology in the office. Imagine pitching to the CEO that you can save a million dollars on the phone bill by getting rid of phones, or a fortune by getting rid of the mailroom.
I laugh at all the conflicting MBA-driven single-budget savings programs I see at work. First you get an email telling everybody to cut down travel and make better use of electronic communications. Then you get another email about how much we're spending on storage and that employees shouldn't store so many files on the network or send them over email - instead you should just meet in-person which is more productive anyway. Big companies are full of millions of MBAs micro-optimizing their department's budget getting bonuses for making small chunks of the bottom line smaller, while watching the top line crash because in the end it means nothing if the company can't actually make products that people want to buy. I've seen MBAs try to shave pennies off of manufacturing processes on products with a 50000% markup and they end up going six months without product on shelves while they try to figure out why the processes are now failing - so to save pennies they end up losing the opportunity to make dollars.
Yeah, but I'd swear that my windows XP work laptop actually takes longer to wake from hibernation than it takes to cold boot. Heaven only knows why. When it is restoring you can count to 3 between each tick on the text-mode progress bar. It was still slow, but not quite so fantastically slow, before they installed full-disk encryption on the thing.
My Chromebook has verified boot and encrypted data and I can boot, login, and reload a half-dozen tabs (including gmail) in less time than it takes for the work PC to respond to the 3-finger salute once the session lock dialog is actually displayed after waking from hibernation. That disk light doesn't go out for about 10 minutes if I do nothing at all. During that time in which the laptop can't be used as anything more than a space heater it probably churns through more data than ever went through the bus in every computer I ever handled from 3rd-12th grade (granted, we're talking pre-pentium days) - and I'd probably include the video RAMDACs in that.
I'm just amazed how much garbage corporate IT manages to load on these things...
Yeah, but short-sighted managers don't see it that way. They figure the employee has to get x amount of work done in a year, and if the computer slows them down then they'll just have to work that many more hours to make it happen so that they don't lose their job in a down economy.
Sure, that improvement might save $8k per employee, but in the mind of a budget-oriented MBA they don't see that $8k unless they can cut bonuses that year. The $8k is largely unmeasurable, and if it can't be measured, it must not be important. Or, so the mindset goes...
That's why privately owned businesses don't have these kinds of problems. The owner walks around at 8AM and sees everybody staring at their computer for 15 minutes doing nothing and even though he can't "measure" it he KNOWS he is losing money. The manager in a fortune 500 company doesn't really care if the company loses money - he cares how much credit he can get for improving some metric that he can sell to his boss for a raise.
Yeah, at work they stopped buying into that logic with the last round of budget cuts. In order to claim a productivity savings you have to name a department and how many people of what job titles will be fired after an IT improvement is made. Saving everybody 7 minutes a day counts as zero savings unless you have 69 employees and plan to fire one of them after buying the SSDs. Then you can go hog wild with them.
I hear you - same experience at work for me. And it isn't any better when restoring from hibernation. The only thing that saves me is having to find/open all the apps I run. I usually start it booting before breakfast so that it is mostly done disk-thrashing by the time I'm done. Then I log into the VPN and then go check my home email on a desktop at home for 10 mins while it does some more thrashing. This is on a two year old laptop running XP with 2GB of RAM - I have computers FAR older at home that boot/login in WAY less time than that, and I even have them running roaming profiles on a samba server.
But, my home PC doesn't have McAfee to the hilt and whatever enterprise-y full disk encryption solution they're using. Those products are sold based on the admin dashboards, not the end-user experience. In fact, half our company was down for a day last year when we had the whole McAfee update-from-hell fiasco.
The other issue is that Chrome doesn't tie compatibility to version numbers. When I update Chrome I don't get a box telling me it is disabling half my extensions/apps. For the most part, everything just works. So, the number is just a number.
Mozilla's problem is that they assume extensions don't work after major version changes, which basically imposes arbitrary breakage. So, the number isn't just a number in their case.
Interesting you mention Gentoo - that is one of the few distros that I'd expect to work fairly well on an i386, since it is built from source.
You'd need to find something to boot off of that contained wget, fdisk, etc, but once that is done you should be able to do just fine. Or, you could just rebuild the install media/etc for i386.
Getting it started would be the only painful part - once you're up and running with an i386 toolchain you should in theory be able to install anything you want. Of course, you'd have to have a pretty minimal install anyway to fit within the RAM constraints on any i386 motherboard.
Well, screening doesn't REQUIRE a body scanner. You could just give the flight crews the TSA pat-down treatment.:) Metal detectors don't use ionizing radiation either. And you can X-ray their bags all you want. They could still get priority in line.
I still say every security company has to hire at least one role player for their Red Teams. They come up with so completely whacked out ways to game the system and subvert anything the "game master" (security officer) can possibly come up with.
The reason this doesn't happen is that in the end the security company needs to sell security, and invariably the customer is never willing to pay for it. So, the security company has to sell them a lot of stuff, make the customer feel more secure, and then they're #1 on the list when the customer has another job.
If you sell somebody a lot of security process/equipment/etc and the result is that your penetration team gets through 40% of the time instead of 90% of the time, the customer is just gong to think that you didn't do your job right. Your competitors are going to do the smart (and dishonest) thing - have the penetration team get through whatever security they previously had (if their company didn't create it in the first place), and then after the improvements are made not get through at all.
When a big company or agency buys "security" the main thing they want is freedom from liability. They don't really care if somebody blows up a plane or steals private customer data or whatever - they just care that they can't get fired since they did due diligence. In the case of government giving extra business to preferred vendors doesn't hurt either.
Here is another problem with not screening pilots - being a pilot doesn't actually mean that you're already able to crash the plane, or at least, not the right plane.
A terrorist wanting to do a 9/11 style attack doesn't just need a plane - they need a big one on a long route carrying lots of fuel and ideally passengers as well. Those routes aren't just assigned to ANY pilots - they're assigned to the most senior pilots. A terrorist isn't likely to make a 25-year career investment to get access to a plane like this.
On the other hand, there are lots of smaller planes on short routes flying out of the same airports as the larger planes. Getting a job to fly one of these planes is much easier. A terrorist would still need to make a decent investment to get a job like this, but we're probably only talking about a few years effort.
So, if flight crews are able to bypass security, then a terrorist could get a job as a pilot of a turboprop, then show up on their day off, bypass security, change clothes in a bathroom, and then board an international flight.
In general, allowing people to bypass security isn't a good thing. If we don't really need all that screening then just get rid of it, and if we do, then make EVERYBODY go through it.
Yup, I'm on TMo and I'm not looking forward to this. I just hope I can keep 4G service on my phone through the end of the contract.
What I'm dreading is having ATT tell me that I'm going to be dropped to edge unless I get a new phone (and possibly re-up my contract). Maybe they'll offer me a free phone, and maybe even without a contract. The problem is that I like my current android phone, and I'm pretty picky about picking out phones. A key feature when selecting my current model was an active mod community and straightforward available rooting mechanism. Somehow I doubt that ATT is going to offer me a phone with those features...
Well, it would make the non-data plans a trivial migration. That is still quite a few phones that could be instantly switched over without much fuss (same frequencies/etc). However, I'd think that any tower hardware capable of serving up 3G data on both the T-mobile and ATT frequencies/protocols could probably handle CDMA/etc just as easily. T-mobile and ATT use very different standards for 3G.
Most likely they'd just run the two networks in parallel for two years and then force everybody on one of them to buy a new phone. That would almost certainly be an ATT-style no-sideloading hard-to-root phone. Ugh - hate to think that I might actually end up switching back to Verizon...
Keep in mind that the lions share of taxes (and population) in the USA come from areas that don't have subway systems, so there isn't a lot of political pressure to invest in them - and lots of pressure to not spend money on them.
From the little I've traveled around Europe I can say that it is VERY different from the US. I tend to travel on business, which gives me the advantage of tending to see how ordinary people live (as opposed to what tourist traps look like). For example, in a typical town in the Netherlands you'll find a small train station or two, with about 5000 bicycles parked outside, and then 6-story buildings for a 5-10 block radius, and then nothing but farmland until you hit the next small town. You can easily walk or bike from the train station to anyplace you'd want to go.
In the USA in an area with similar average population density you wouldn't find any buildings taller than 2 stories (or maybe the rare 3), and each building would be about 200 meters apart at least, surrounded by a large parking lot. They'd cover the same area as the space between two European towns. So, instead of one bustling town that you can walk around, you have this general smear of population everywhere. Within that 5-10 "block" radius of a hypothetical train station you'd find fewer people working than you'd find within 1 block in a European town. Of course, there is no real train station, since there would be no point in having one.
It is just a very different culture. In the USA just about everybody has a quarter acre of lawn outside their home (on average). This is considered of higher value than being able to walk to get groceries.
And having traveled roads in a number of countries, I'd say the urban access roads in the USA are better than what you'll find just about anywhere. For starters, virtually all the roads all actually have sidewalks and separate lanes for opposing traffic.:) The highway system is better in some European nations, but not by a great deal, and the US highways clearly carry far more cars. I think the US just has different priorities.
All that said, I'd love to see practical public transit options in the US. However, so far they just don't exist for the rural population densities. Maybe when cars are able to drive themselves that will solve the problem (cars travelling in formation operating on a fee-for-trip model would probably be about as cost-effective and environmentally sound as trains are).
That is a LOT of cash. I'm surprised most linux distros don't negotiate their own deals with Google. They could charge Google to keep their default browser setting in Firefox (or for that matter Chrome), for example, or charge Bing to switch it. Granted, the volume for linux will be lower than for windows, but I imagine that quite a few people still install firefox/chrome/etc.
Now, the trademarks could become an issue. A way around that is to patch in the change post-install so that the distro doesn't actually distribute a modified version of the browser. Or, they could just iceweasel it.:)
So, let's figure that out. 250MW of power, on a motorcycle+passenger weighing 500kg (it has to be a pretty heavy bike if the engine itself weighs 227kg I figure).
In one second your bike has 250MJ of kinetic energy, which works out to a speed of 1582 mph. That "only" works out to 72g of acceleration.
But, yes, falls somewhere between a carrier takeoff and an artillery round.
I think firing is not likely to be a big problem, so much as hiring in the first place.
You send out a resume to 100 companies, and you don't get any requests for interviews. Is it because they just don't need you or have much better resumes, or is because you have photos from Comic-Con on your facebook page? You'll never be told.
Or, suppose your boss is told they need to lay off one person. You're the guy who gets let go. Did facebook have anything to do with it? All you'd be told is that your services are no longer required and best of luck.
Re:Experiments performed only on 3 test subjects
on
Cancer Cured By HIV
·
· Score: 1
Think of it this way:
I'd like to test whether taping copper pennies to ones feet can cure cancer. I'd like to have a million dollars to do controlled randomized drug trials. If successful I could revolutionize the practice of medicine for literally a cost of pennies.
Do you think I'm likely to get funded, even though the treatment is clearly "high impact?" There are all kinds of ideas out there like this one, and none have worked yet. This one might or might not actually pan out, though it clearly would be wonderful if it did.
I don't think they were snubbed - they just weren't lucky. Stab-in-the-dark research doesn't tend to get a lot of funding, except perhaps in start-ups.
Re:It'll never make it through FDA trials
on
Cancer Cured By HIV
·
· Score: 1
Why would Big Pharma cure cancer if they could convert it from a deadly disease into a chronic condition (and profit from selling you the treatment) instead?
Big Pharma doesn't have a working chronic treatment for cancer, so you're asking why would a big company choose to make a boatload of money now, instead of making no money now but maybe hypothetically in the future they could make less money for a longer time.
The answer is simple - the executive gets his bonus based on this quarter's sales, not the product sales ten years from now. Sure, all things being equal the company is going to pursue the most profitable product, but they're not going to avoid making money if the opportunity is there.
This is blue-sky research. Blue sky research is important because it has the potential to change everything, even though most of the time it is just a money sink. A typical company isn't going to invest much in blue sky research, since it could instead invest the money in something more likely to lead to a near-term return. So, big pharma isn't likely to discover the cure for cancer for the same reason that they would be the first to market one if it were found - short-sightedness.
Sorry, to bring my example full-circle, consider this situation:
1. You deploy your evolutionary anti-virus system into production. 2. Your postfix server gets stuck in a loop and starts churning out spam.
Would the antivirus be likely to pick up on the postfix problem and kill any misbehaving server processes? Even though the algorithm might pick up any random program that is behaving badly fairly easily, it could miss the postfix problem, since it is a fault in a process that it was trained to recognize as being valid. Detection code that stood a chance of targeting postfix would have been wiped out during training, and so unless the change is huge it could get missed even by millions of different detection algorithms.
This is the situation that often arises with cancer - a cancer cell is just too similar to a normal cell for the immune system to pick up on it much of the time. As with the anti-virus program, the system has been trained to ignore it. On the other hand, somebody could go in and create a detection algorithm specifically engineered to spot misbehaving postfix services and splice it into the random region of your antivirus program.
T cell receptors (the things that stick out of T cells which allow them to detect their prey) are incredible biologically because the body makes them up at semi-random when generating new T cells; it does the same for antibodies. However, we only have so many building blocks to choose from when making them, and the receptors we need to target leukaemia aren't possible.
Why can't naturally-occurring T-cell receptors target leukemia? Is it because it requires some sequence that isn't possible starting from the genes immune cells work with? Or, is it a result of anti-selection during development?
I'm not an immunologist, but my understanding is that the random processes that lead to the body being able to generate antibodies against a huge number of foreign substances that it has never encountered can also lead to the body being able to generate antibodies that would lead to the destruction of its own cells. However, during early development immune cells that detect anything kill themselves, since the assumption is that the only stuff around is the embryo's own tissue. By the time a child is born, the only immune cells left contain random receptors that DON'T have an affinity for anything alive.
Ok, how about a computer analogy for slashdot. Imagine you want to build a heuristic antivirus system, that works against future unknown threats AND threat mechanisms. Instead of having one software package that tries to figure out what is bad, you instead have a software package that has a detection algorithm that consists completely of random bytes, and you create so many variants that you have hundreds of millions that actually execute. Then you run each scanner against your known-good system, and destroy any program whose detection algorithm is triggered. Then you deploy what is left for production use - any behavior that is unusual stands a chance of triggering one of the millions of antivirus programs that is running. Attempting to defeat this mechanism would be very difficult, since those programs could be looking for almost anything. The risk of course is false positives, and that happens in the human immune system also (causing stuff like allergies, poison ivy rashes, or nasty stuff like lupus).
The immune system is basically a big example of an evolutionary algorithm, and the algorithm itself is of course tuned by evolution (so, should we call it a second degree evolutionary algorithm?). There are actually many aspects of the immune system that work in this way, and not just how antibodies and receptors are formed.
Slashdot Mirror
The FDA does regulate "adverse event reporting." Look at any bottle of drugs you have (in the USA). Somewhere on it is a phone number. If you call that number and report that something bad happened after you took that pill you'll kick off a complex processes designed to ensure that all the details of your problem are captured, tracked, and ultimately reported back to the FDA. The FDA can require companies to study any trends that arise, and drugs actually have been taken off the market as a result.
The concern is that if you post "this drug made me sick" on a facebook wall, you just informed the company about an adverse event that the company is now legally required to follow-up on. However, to do this they might have to subpoena your personal information from Facebook, your ISP, and who-knows how many other intermediaries if you tried to make the account anonymous. They then need to try to track you and your doctor down and find out what happened. This is a legal nightmare. If you just called the number on your prescription bottle they'd gather the legally-required info up-front, and would know how to track you down.
I'd be a fan of just exempting social media from these kinds of requirements - if you want to moan about getting a stomachache on facebook have fun, but you shouldn't expect it to result in an investigation. However, government agencies aren't known for common sense, and they are known for beating people up over subtle interpretations of the rules.
However, I'd think that any tower hardware capable of serving up 3G data on both the T-mobile and ATT frequencies/protocols could probably handle CDMA/etc just as easily.
And you'd be wrong. At least not without spending assloads of money.
Can you supply details? Does ATT or T-Mo have lots of hardware deployed which could be re-tasked to serve the other networks 3G phones but not a CDMA-based network? My understanding is that the ATT and T-Mo 3G networks are fairly different (different frequencies and protocols), so the only way I figure the same hardware could support both is if it uses software-defined radio and has very wide multiband support, and hardware with these capabilities probably could handle CDMA. Then again, I could be wrong on that. And, I wouldn't be surprised if none of the towers out there could support both without hardware changes.
So, my employer does this kind of nonsense, and they have literally billions of dollars in "cash" on the balance sheet. So, the only cash flow issue is making the cash flow look good - not that they can't write checks.
And the MBAs have learned a long time ago that the goal isn't to make the company do well - the goal is to make your paycheck as large as possible. That does not require making the company do well - it merely requires convincing the boss that you're making the company do well, usually by reducing the size of the budget under your immediate control.
Yup, but the problem with your argument is that the kinds of savings you're arguing for are hard to measure. Sure, you can measure PC boot time, but what is the actual production output of a department after you fix their boot times? Chances are it would be hard to trace your improvement to real-world impact. I'm sure it is there and it pays for itself, but if you can't show it in numbers, then you can't get a bonus for implementing it. You can, however, get in trouble if you mess things up when implementing it. So, to an IT manager what you propose is all downside.
The whole IT-is-a-cost-center MBA argument would be recognized as being silly if applied to any other less-complex technology in the office. Imagine pitching to the CEO that you can save a million dollars on the phone bill by getting rid of phones, or a fortune by getting rid of the mailroom.
I laugh at all the conflicting MBA-driven single-budget savings programs I see at work. First you get an email telling everybody to cut down travel and make better use of electronic communications. Then you get another email about how much we're spending on storage and that employees shouldn't store so many files on the network or send them over email - instead you should just meet in-person which is more productive anyway. Big companies are full of millions of MBAs micro-optimizing their department's budget getting bonuses for making small chunks of the bottom line smaller, while watching the top line crash because in the end it means nothing if the company can't actually make products that people want to buy. I've seen MBAs try to shave pennies off of manufacturing processes on products with a 50000% markup and they end up going six months without product on shelves while they try to figure out why the processes are now failing - so to save pennies they end up losing the opportunity to make dollars.
Yeah, but I'd swear that my windows XP work laptop actually takes longer to wake from hibernation than it takes to cold boot. Heaven only knows why. When it is restoring you can count to 3 between each tick on the text-mode progress bar. It was still slow, but not quite so fantastically slow, before they installed full-disk encryption on the thing.
My Chromebook has verified boot and encrypted data and I can boot, login, and reload a half-dozen tabs (including gmail) in less time than it takes for the work PC to respond to the 3-finger salute once the session lock dialog is actually displayed after waking from hibernation. That disk light doesn't go out for about 10 minutes if I do nothing at all. During that time in which the laptop can't be used as anything more than a space heater it probably churns through more data than ever went through the bus in every computer I ever handled from 3rd-12th grade (granted, we're talking pre-pentium days) - and I'd probably include the video RAMDACs in that.
I'm just amazed how much garbage corporate IT manages to load on these things...
Yeah, but short-sighted managers don't see it that way. They figure the employee has to get x amount of work done in a year, and if the computer slows them down then they'll just have to work that many more hours to make it happen so that they don't lose their job in a down economy.
Sure, that improvement might save $8k per employee, but in the mind of a budget-oriented MBA they don't see that $8k unless they can cut bonuses that year. The $8k is largely unmeasurable, and if it can't be measured, it must not be important. Or, so the mindset goes...
That's why privately owned businesses don't have these kinds of problems. The owner walks around at 8AM and sees everybody staring at their computer for 15 minutes doing nothing and even though he can't "measure" it he KNOWS he is losing money. The manager in a fortune 500 company doesn't really care if the company loses money - he cares how much credit he can get for improving some metric that he can sell to his boss for a raise.
Yeah, at work they stopped buying into that logic with the last round of budget cuts. In order to claim a productivity savings you have to name a department and how many people of what job titles will be fired after an IT improvement is made. Saving everybody 7 minutes a day counts as zero savings unless you have 69 employees and plan to fire one of them after buying the SSDs. Then you can go hog wild with them.
Gotta love the MBAs.
I hear you - same experience at work for me. And it isn't any better when restoring from hibernation. The only thing that saves me is having to find/open all the apps I run. I usually start it booting before breakfast so that it is mostly done disk-thrashing by the time I'm done. Then I log into the VPN and then go check my home email on a desktop at home for 10 mins while it does some more thrashing. This is on a two year old laptop running XP with 2GB of RAM - I have computers FAR older at home that boot/login in WAY less time than that, and I even have them running roaming profiles on a samba server.
But, my home PC doesn't have McAfee to the hilt and whatever enterprise-y full disk encryption solution they're using. Those products are sold based on the admin dashboards, not the end-user experience. In fact, half our company was down for a day last year when we had the whole McAfee update-from-hell fiasco.
The other issue is that Chrome doesn't tie compatibility to version numbers. When I update Chrome I don't get a box telling me it is disabling half my extensions/apps. For the most part, everything just works. So, the number is just a number.
Mozilla's problem is that they assume extensions don't work after major version changes, which basically imposes arbitrary breakage. So, the number isn't just a number in their case.
Interesting you mention Gentoo - that is one of the few distros that I'd expect to work fairly well on an i386, since it is built from source.
You'd need to find something to boot off of that contained wget, fdisk, etc, but once that is done you should be able to do just fine. Or, you could just rebuild the install media/etc for i386.
Getting it started would be the only painful part - once you're up and running with an i386 toolchain you should in theory be able to install anything you want. Of course, you'd have to have a pretty minimal install anyway to fit within the RAM constraints on any i386 motherboard.
Well, screening doesn't REQUIRE a body scanner. You could just give the flight crews the TSA pat-down treatment. :) Metal detectors don't use ionizing radiation either. And you can X-ray their bags all you want. They could still get priority in line.
I still say every security company has to hire at least one role player for their Red Teams. They come up with so completely whacked out ways to game the system and subvert anything the "game master" (security officer) can possibly come up with.
The reason this doesn't happen is that in the end the security company needs to sell security, and invariably the customer is never willing to pay for it. So, the security company has to sell them a lot of stuff, make the customer feel more secure, and then they're #1 on the list when the customer has another job.
If you sell somebody a lot of security process/equipment/etc and the result is that your penetration team gets through 40% of the time instead of 90% of the time, the customer is just gong to think that you didn't do your job right. Your competitors are going to do the smart (and dishonest) thing - have the penetration team get through whatever security they previously had (if their company didn't create it in the first place), and then after the improvements are made not get through at all.
When a big company or agency buys "security" the main thing they want is freedom from liability. They don't really care if somebody blows up a plane or steals private customer data or whatever - they just care that they can't get fired since they did due diligence. In the case of government giving extra business to preferred vendors doesn't hurt either.
Here is another problem with not screening pilots - being a pilot doesn't actually mean that you're already able to crash the plane, or at least, not the right plane.
A terrorist wanting to do a 9/11 style attack doesn't just need a plane - they need a big one on a long route carrying lots of fuel and ideally passengers as well. Those routes aren't just assigned to ANY pilots - they're assigned to the most senior pilots. A terrorist isn't likely to make a 25-year career investment to get access to a plane like this.
On the other hand, there are lots of smaller planes on short routes flying out of the same airports as the larger planes. Getting a job to fly one of these planes is much easier. A terrorist would still need to make a decent investment to get a job like this, but we're probably only talking about a few years effort.
So, if flight crews are able to bypass security, then a terrorist could get a job as a pilot of a turboprop, then show up on their day off, bypass security, change clothes in a bathroom, and then board an international flight.
In general, allowing people to bypass security isn't a good thing. If we don't really need all that screening then just get rid of it, and if we do, then make EVERYBODY go through it.
Yup, I'm on TMo and I'm not looking forward to this. I just hope I can keep 4G service on my phone through the end of the contract.
What I'm dreading is having ATT tell me that I'm going to be dropped to edge unless I get a new phone (and possibly re-up my contract). Maybe they'll offer me a free phone, and maybe even without a contract. The problem is that I like my current android phone, and I'm pretty picky about picking out phones. A key feature when selecting my current model was an active mod community and straightforward available rooting mechanism. Somehow I doubt that ATT is going to offer me a phone with those features...
Well, it would make the non-data plans a trivial migration. That is still quite a few phones that could be instantly switched over without much fuss (same frequencies/etc). However, I'd think that any tower hardware capable of serving up 3G data on both the T-mobile and ATT frequencies/protocols could probably handle CDMA/etc just as easily. T-mobile and ATT use very different standards for 3G.
Most likely they'd just run the two networks in parallel for two years and then force everybody on one of them to buy a new phone. That would almost certainly be an ATT-style no-sideloading hard-to-root phone. Ugh - hate to think that I might actually end up switching back to Verizon...
Keep in mind that the lions share of taxes (and population) in the USA come from areas that don't have subway systems, so there isn't a lot of political pressure to invest in them - and lots of pressure to not spend money on them.
From the little I've traveled around Europe I can say that it is VERY different from the US. I tend to travel on business, which gives me the advantage of tending to see how ordinary people live (as opposed to what tourist traps look like). For example, in a typical town in the Netherlands you'll find a small train station or two, with about 5000 bicycles parked outside, and then 6-story buildings for a 5-10 block radius, and then nothing but farmland until you hit the next small town. You can easily walk or bike from the train station to anyplace you'd want to go.
In the USA in an area with similar average population density you wouldn't find any buildings taller than 2 stories (or maybe the rare 3), and each building would be about 200 meters apart at least, surrounded by a large parking lot. They'd cover the same area as the space between two European towns. So, instead of one bustling town that you can walk around, you have this general smear of population everywhere. Within that 5-10 "block" radius of a hypothetical train station you'd find fewer people working than you'd find within 1 block in a European town. Of course, there is no real train station, since there would be no point in having one.
It is just a very different culture. In the USA just about everybody has a quarter acre of lawn outside their home (on average). This is considered of higher value than being able to walk to get groceries.
And having traveled roads in a number of countries, I'd say the urban access roads in the USA are better than what you'll find just about anywhere. For starters, virtually all the roads all actually have sidewalks and separate lanes for opposing traffic. :) The highway system is better in some European nations, but not by a great deal, and the US highways clearly carry far more cars. I think the US just has different priorities.
All that said, I'd love to see practical public transit options in the US. However, so far they just don't exist for the rural population densities. Maybe when cars are able to drive themselves that will solve the problem (cars travelling in formation operating on a fee-for-trip model would probably be about as cost-effective and environmentally sound as trains are).
That is a LOT of cash. I'm surprised most linux distros don't negotiate their own deals with Google. They could charge Google to keep their default browser setting in Firefox (or for that matter Chrome), for example, or charge Bing to switch it. Granted, the volume for linux will be lower than for windows, but I imagine that quite a few people still install firefox/chrome/etc.
Now, the trademarks could become an issue. A way around that is to patch in the change post-install so that the distro doesn't actually distribute a modified version of the browser. Or, they could just iceweasel it. :)
So, let's figure that out. 250MW of power, on a motorcycle+passenger weighing 500kg (it has to be a pretty heavy bike if the engine itself weighs 227kg I figure).
In one second your bike has 250MJ of kinetic energy, which works out to a speed of 1582 mph. That "only" works out to 72g of acceleration.
But, yes, falls somewhere between a carrier takeoff and an artillery round.
I think firing is not likely to be a big problem, so much as hiring in the first place.
You send out a resume to 100 companies, and you don't get any requests for interviews. Is it because they just don't need you or have much better resumes, or is because you have photos from Comic-Con on your facebook page? You'll never be told.
Or, suppose your boss is told they need to lay off one person. You're the guy who gets let go. Did facebook have anything to do with it? All you'd be told is that your services are no longer required and best of luck.
Read his last sentence more carefully. :)
Think of it this way:
I'd like to test whether taping copper pennies to ones feet can cure cancer. I'd like to have a million dollars to do controlled randomized drug trials. If successful I could revolutionize the practice of medicine for literally a cost of pennies.
Do you think I'm likely to get funded, even though the treatment is clearly "high impact?" There are all kinds of ideas out there like this one, and none have worked yet. This one might or might not actually pan out, though it clearly would be wonderful if it did.
I don't think they were snubbed - they just weren't lucky. Stab-in-the-dark research doesn't tend to get a lot of funding, except perhaps in start-ups.
Why would Big Pharma cure cancer if they could convert it from a deadly disease into a chronic condition (and profit from selling you the treatment) instead?
Big Pharma doesn't have a working chronic treatment for cancer, so you're asking why would a big company choose to make a boatload of money now, instead of making no money now but maybe hypothetically in the future they could make less money for a longer time.
The answer is simple - the executive gets his bonus based on this quarter's sales, not the product sales ten years from now. Sure, all things being equal the company is going to pursue the most profitable product, but they're not going to avoid making money if the opportunity is there.
This is blue-sky research. Blue sky research is important because it has the potential to change everything, even though most of the time it is just a money sink. A typical company isn't going to invest much in blue sky research, since it could instead invest the money in something more likely to lead to a near-term return. So, big pharma isn't likely to discover the cure for cancer for the same reason that they would be the first to market one if it were found - short-sightedness.
Sorry, to bring my example full-circle, consider this situation:
1. You deploy your evolutionary anti-virus system into production.
2. Your postfix server gets stuck in a loop and starts churning out spam.
Would the antivirus be likely to pick up on the postfix problem and kill any misbehaving server processes? Even though the algorithm might pick up any random program that is behaving badly fairly easily, it could miss the postfix problem, since it is a fault in a process that it was trained to recognize as being valid. Detection code that stood a chance of targeting postfix would have been wiped out during training, and so unless the change is huge it could get missed even by millions of different detection algorithms.
This is the situation that often arises with cancer - a cancer cell is just too similar to a normal cell for the immune system to pick up on it much of the time. As with the anti-virus program, the system has been trained to ignore it. On the other hand, somebody could go in and create a detection algorithm specifically engineered to spot misbehaving postfix services and splice it into the random region of your antivirus program.
T cell receptors (the things that stick out of T cells which allow them to detect their prey) are incredible biologically because the body makes them up at semi-random when generating new T cells; it does the same for antibodies. However, we only have so many building blocks to choose from when making them, and the receptors we need to target leukaemia aren't possible.
Why can't naturally-occurring T-cell receptors target leukemia? Is it because it requires some sequence that isn't possible starting from the genes immune cells work with? Or, is it a result of anti-selection during development?
I'm not an immunologist, but my understanding is that the random processes that lead to the body being able to generate antibodies against a huge number of foreign substances that it has never encountered can also lead to the body being able to generate antibodies that would lead to the destruction of its own cells. However, during early development immune cells that detect anything kill themselves, since the assumption is that the only stuff around is the embryo's own tissue. By the time a child is born, the only immune cells left contain random receptors that DON'T have an affinity for anything alive.
Ok, how about a computer analogy for slashdot. Imagine you want to build a heuristic antivirus system, that works against future unknown threats AND threat mechanisms. Instead of having one software package that tries to figure out what is bad, you instead have a software package that has a detection algorithm that consists completely of random bytes, and you create so many variants that you have hundreds of millions that actually execute. Then you run each scanner against your known-good system, and destroy any program whose detection algorithm is triggered. Then you deploy what is left for production use - any behavior that is unusual stands a chance of triggering one of the millions of antivirus programs that is running. Attempting to defeat this mechanism would be very difficult, since those programs could be looking for almost anything. The risk of course is false positives, and that happens in the human immune system also (causing stuff like allergies, poison ivy rashes, or nasty stuff like lupus).
The immune system is basically a big example of an evolutionary algorithm, and the algorithm itself is of course tuned by evolution (so, should we call it a second degree evolutionary algorithm?). There are actually many aspects of the immune system that work in this way, and not just how antibodies and receptors are formed.
Money. It's the cost of the Lotus Notes licenses that's preventing us from deploying killbots.
Couldn't they just grab a copy of MySQL, Eudora and Edlin and call it a day?
Definitely not, that would be exceeding the specifications.