Sorry, but the magtek "solution" is a band-aid at best, and far more likely to be snake oil. It's expensive to deploy the fancy proprietary hardware to every single merchant, and as soon as the cloners improve their technology the whole thing fails epically.
The entire "security" of the magtek system comes from a technical difficulty that nobody's had the economic incentive to try to break, not that it's technically unbreakable.
If adopted, I predict that magtek will make their money, then collapse under the weight of the inevitable lawsuits.
Joe Crackhead might not be able to do much on his own, but an organized crime ring can use people like Joe to deliver a stream of stolen cards. If a smart crook can find a way to exploit them in batches, Joe will continue to steal them as long as he gets paid.
Regardless, the risks to ordinary people still drop by an order of magnitude or four.
Digital photos turn out to be one of those data types where a GUI excels. If I knew from its file name that IMG_3047.JPG was the east face of Mt. Fuji, I wouldn't need the categorizations. Similarly, since I have to already look at the image itself in a GUI tool to identify it, I wouldn't want to retype the file name in order to add a category if I could simply select it.
The other drawback to your suggestion is it's still essentially an "external database" approach to tagging. Sure, the "database" is the file system, but it's no longer intrinsic to the image files, which is what the OP was requesting. If I copy ~/photos/mt_fuji/east_face/IMG_3047.JPG to the root of a flash stick, I've lost all that useful information. Sure, that's a problem with every external database solution, but it's not what the guy wants.
It might be cool to have a program that automates what you describe. It's probably a SMoP (Small Matter of Python) to write up code that creates a directory structure based on the EXIF tags, and symbolically links all the pictures to directories according to something like your scheme above. Turn the program loose on a pile of tagged digital photos, and then you can script your heart out. You could probably do the reverse, too, and tag photos with the components of their folder names (not sure how useful that would be.)
This license is not written very clearly, but that's what it means.
And that's what good lawyers are for. Proving that the terms and conditions are unintelligible or unusable as published could exempt you from being compelled to follow them. Was there a shrink-wrap license on the camcorder, or some adhesive tape on the battery door announcing "by breaking this seal you are agreeing to the legal terms and conditions printed on page 734 of the instruction book?" If so, were those instructions and obligations clear? Did the outside of the box say "inside this box are heinous terms and conditions that will cost you millions of dollars in case you shoot the next YouTube phenomenon?"
I think if someone shot the Ultimate Viral Video in MPEG-LA and the MAFIAA decided to go after him for $0.02/download, good legal representation would probably successfully challenge the contract; with luck, the rest of the world would find out about how to defend against it as well, rendering the piece of sh!t pretty much useless.
Shotwell does not do what the OP is asking for. It does not modify the EXIF tags on the pictures. It's specifically not"File-Centric" as it stores your edits, crops, cuts, etc., only in its database. It renders all the change info on the fly when you display the photos.
But it is kind of a nice tool. Mostly user friendly, and since it's still version 0.5, I suspect it will receive a lot of tweaks and improvements before it's marked "stable".
It would also make being a pilot a more bearable because a pilot could simple hand over the controls to another pilot after an 8-hour shift and go home to his family. That would get rid of things like pilot fatigue and allow them to have normal circadian rhythms in tune with their local time zone.
You're ignoring the very real threat posed by big, stupid corporations. Some pencil pushers will overbook pilots, or take whatever cost savings measures they can. Sure, they'll promise a certain level of staffing, but that will be the first thing forgotten in the next bankruptcy or merger.
Yes, the administrators are requiring it, but likely at the behest of the technicians. And the techs are enforcing it by not performing unless the order says "please". Kind of makes a mockery of the term "order" there, too.
I imagine this is going on today.
Original order: "Draw Mr. Smith's blood." Technician: "Denied, you didn't write the magic word." Revised order: "Draw Mr. Smith's blood by 9:00, and if you ever question my orders again I'll have your arse sacked." Technician: "Those are magic words. Here's your lab results."
Just an observation: if you trust that the existing wear leveling algorithm has been performing properly, you can safely reset its state as well. Think about it. Resetting them won't unbalance it by more than one extra write or so for the affected blocks.
Flash bits typically don't fail from reading, they fail from the stress of writing. And they don't wear out at exactly the 100,000th write, so you never know precisely which bit is going to be "more susceptible" to failure than another. So as long as the wear leveling algorithms work every time you write to the chip, resetting the state occasionally isn't going to impact the lifetime of the chip by significantly overwriting one block more than others. Resetting the state on a minute-by-minute basis, however, is likely to overwork the state bits first, and possibly weaken the block at the algorithm's "first block allocated from a reset state".
Your only risk is overwriting the leveling state would also reset whatever counters might signal the controller that "this is my 150,000th write, my warranty is about to expire and the chances of failure have now risen to over.000002% per write operation." I know S.M.A.R.T. monitors work with those kinds of statistics for spinning drives, but I've never bothered to look at if they report on flash drive statistics.
I don't think extra controls would have make a whit of difference in the validity of this particular study. There's simply too much variance in the environment to draw a conclusion. Valid results would be drowned in the noise of other potential natural causes of CCD: toxins, pesticides, fungi, viral infections, feral bees, predators, parasites, etc.
A proper experiment would require hundreds if not thousands of hives. And yes, they'd need better protocols for handling the phones (I like your theory about outgassing), although the idea of having the phones actually inside the hives is still absurd as a potential cause of CCD, as this experiment likely provided the world with the only two hives ever to actually have cell phones inside them.
Nope. All the bees in a single field could be impacted by a nearby external uncontrolled stimulus: anything from an early frost, drought, pesticide applications, diseases, mites, fungi, competing colonies of nearby bees, or even dirt on the shoes of the guy changing the cell phone batteries. You'd need a really large set of samples to figure out if there was even a measurable impact by using a Faraday cage. I seriously doubt that a hundred fields would provide enough samples.
Four samples would yield nothing more significant that the current article, which is another way of saying "worthless".
the iPhone threw out the gimped phone device business model crammed down our throats by the phone companies.
"gimped phone device"? Holy shit, the iPhone is the most gimped phone I've had in 10 years. The lock-in to the app store? A Bluetooth stack that doesn't support OBEX or AVRCP? No J2ME?
People actually celebrated when the black turtleneckers deigned to bestow MMS upon their undeserving countenance. That's how bad Apple is, and yet you've bought the propaganda hook, line and sinker. Wake up.
I wasn't even sliding into the realm of actual criminal corporate behavior. Just working within the existing system, every activity and idea is turned into dollars before being discussed. I don't know that you can change that at any meaningful level, as that is how success is measured in the financial world. Everything is dollars (well, until the saying becomes "everything is yuan", anyway.)
I'm saying that our rights are undervalued in the current system, and the only way to change it is for us to assign them a lot more value through the act of lawsuits. If we don't sue then our rights are actually being fairly valued within the system, and they will continue to be shat upon by people like Tim O'Reilly as there is no financial penalty for trying.
Why not artificial meat in the garage? It's pretty much cloning tissue, which is possible at small lab scales. The hard part is figuring out what to clone, and how to grow it. That means lots of experiments in environmental conditions, cultural media, etc. Plenty of room for amateurs to try things, from inventing cheap yet highly stable heating systems to efficient nutrient sources.
You've hit on kind of the sweet spot there. I agree with you that the scientific world seems to have had many of its boundaries pushed beyond the capacity of the average home experimenter, but the artistic world has no such boundaries. Fun and artistic electromechanical toys and hacks are still novel. Look at shows like Burning Man, sites like hackaday, magazines like Make:. They're filled with people interested in the act of creation. And last night my brother-in-law introduced me to Farm Show magazine (farmshow.com) which is a compendium of hacks and homebuilt machines that farmers have created out of necessity and imagination. It has a lot of really cool homemade things in it.
And if you're looking to monetize it, handmade and homemade mechanical equipment has a very visceral appeal to a lot of people. The potential to sell a unique device is high. And you can get involved for any amount of money, from repurposing junk bits from broken VCRs to building a nicely equipped machine shop.
My point is that corporations will continue to assess personal privacy issues as low-dollar-risk line items, even if a few executives go to jail. There's personal incentive for one or two people, but not the corporation as a whole. "Let's continue to screw with people's privacy but let's set up a scapegoat to take the fall" then becomes the unwritten corporate strategy. A jailed CTO won't result in a multimillion dollar hit to the shareholders, as much as the CTO's ego wishes it to be true.
But if analysts and investors are convinced that messing with privacy issues makes stock worthless, they will stop funding companies who try to violate them. Every stinking decision on the Street is based on the almighty dollar.
Small companies would retain the equal opportunity to be sued into oblivion. They'd just be forgotten faster than an operation the size of Facebook.
Just one effective lawsuit against a mom'n'pop shop would likely result in the personal bankruptcy of the owners. That's never a pleasant thought to a small business owner, which is why I'd trust them to have better intentions to do right by their customers. Of course as a mom'n'pop, I'd also not be surprised by a weaker implementation of security measures.
Sorry, but that won't work. As a typical amoral shareholder or mutual funds investor, I'm perfectly OK with the CEO going to jail as long as my dividend checks keep arriving in the mailbox. The market runs only on greed, not fear of incarceration of "other people".
But imagine what would happen if Facebook was sued out of existence because of this change to their privacy policy. The next company to talk about loosening their privacy policy would see their share value dropping in half, as the wary investors divest as fast as they can.
What needs to happen is that a lot of people who were adversely impacted by this have to file giant lawsuits. Let's say that ten thousand people lost their jobs by having "drunk college pictures" revealed to their employers. If I were one of them, I'd sue for lost wages on a career where I would have potentially made $5,000,000 over my lifetime. Get a thousand victims to file those lawsuits, and the company collapses under the burden. Message to corporations: if you mess with our privacy, we will parade your rotting corpse down Wall Street and toss it in the Hudson River.
This is still America. We have the right to file lawsuits if we're aggrieved. It's the only mechanism by which individuals can teach corporations lessons that will stick.
It's become the fashion to lump everything together, as if performances, images, tangible goods, rights, efforts, and ideas are all exactly the same kinds of assets and should be treated exactly the same by corporations, governments, and individuals. That's happened because business students are taught to convert everything to dollars, assign a value to risk, and then simply slide the numbers around on an Excel spreadsheet until the biggest one pops out at the bottom.
The problem is that the dollar value they assign to risk is based on the imaginations of some not-very-creative people, and is only the risk to them, not to the end users. "Well, if we screw with the privacy policy, our risk is that we'll lose less than 0.5% of our users. That's equal to ad revenue of $3,000,000. The ROI on increased ads is projected to be $10,000,000. This change will pay itself back in months, so just do it."
What really has to happen is truckloads and truckloads of lawsuits have to be filed against them, by people whose privacy was violated. They have to learn that if they mess with our privacy, it will cost them billions of dollars in settlements and legal fees -- to the brink of bankruptcy, and even over the edge. That is the only time corporations will start respecting our rights -- when violating them is guaranteed to flush their bottom lines into the toilet.
Slashdot Mirror
Sorry, but the magtek "solution" is a band-aid at best, and far more likely to be snake oil. It's expensive to deploy the fancy proprietary hardware to every single merchant, and as soon as the cloners improve their technology the whole thing fails epically.
The entire "security" of the magtek system comes from a technical difficulty that nobody's had the economic incentive to try to break, not that it's technically unbreakable.
If adopted, I predict that magtek will make their money, then collapse under the weight of the inevitable lawsuits.
Joe Crackhead might not be able to do much on his own, but an organized crime ring can use people like Joe to deliver a stream of stolen cards. If a smart crook can find a way to exploit them in batches, Joe will continue to steal them as long as he gets paid.
Regardless, the risks to ordinary people still drop by an order of magnitude or four.
Digital photos turn out to be one of those data types where a GUI excels. If I knew from its file name that IMG_3047.JPG was the east face of Mt. Fuji, I wouldn't need the categorizations. Similarly, since I have to already look at the image itself in a GUI tool to identify it, I wouldn't want to retype the file name in order to add a category if I could simply select it.
The other drawback to your suggestion is it's still essentially an "external database" approach to tagging. Sure, the "database" is the file system, but it's no longer intrinsic to the image files, which is what the OP was requesting. If I copy ~/photos/mt_fuji/east_face/IMG_3047.JPG to the root of a flash stick, I've lost all that useful information. Sure, that's a problem with every external database solution, but it's not what the guy wants.
It might be cool to have a program that automates what you describe. It's probably a SMoP (Small Matter of Python) to write up code that creates a directory structure based on the EXIF tags, and symbolically links all the pictures to directories according to something like your scheme above. Turn the program loose on a pile of tagged digital photos, and then you can script your heart out. You could probably do the reverse, too, and tag photos with the components of their folder names (not sure how useful that would be.)
This license is not written very clearly, but that's what it means.
And that's what good lawyers are for. Proving that the terms and conditions are unintelligible or unusable as published could exempt you from being compelled to follow them. Was there a shrink-wrap license on the camcorder, or some adhesive tape on the battery door announcing "by breaking this seal you are agreeing to the legal terms and conditions printed on page 734 of the instruction book?" If so, were those instructions and obligations clear? Did the outside of the box say "inside this box are heinous terms and conditions that will cost you millions of dollars in case you shoot the next YouTube phenomenon?"
I think if someone shot the Ultimate Viral Video in MPEG-LA and the MAFIAA decided to go after him for $0.02/download, good legal representation would probably successfully challenge the contract; with luck, the rest of the world would find out about how to defend against it as well, rendering the piece of sh!t pretty much useless.
Shotwell does not do what the OP is asking for. It does not modify the EXIF tags on the pictures. It's specifically not"File-Centric" as it stores your edits, crops, cuts, etc., only in its database. It renders all the change info on the fly when you display the photos.
But it is kind of a nice tool. Mostly user friendly, and since it's still version 0.5, I suspect it will receive a lot of tweaks and improvements before it's marked "stable".
It would also make being a pilot a more bearable because a pilot could simple hand over the controls to another pilot after an 8-hour shift and go home to his family. That would get rid of things like pilot fatigue and allow them to have normal circadian rhythms in tune with their local time zone.
You're ignoring the very real threat posed by big, stupid corporations. Some pencil pushers will overbook pilots, or take whatever cost savings measures they can. Sure, they'll promise a certain level of staffing, but that will be the first thing forgotten in the next bankruptcy or merger.
mushrooms!
mushrooms
Yes, the administrators are requiring it, but likely at the behest of the technicians. And the techs are enforcing it by not performing unless the order says "please". Kind of makes a mockery of the term "order" there, too.
I imagine this is going on today.
Original order: "Draw Mr. Smith's blood."
Technician: "Denied, you didn't write the magic word."
Revised order: "Draw Mr. Smith's blood by 9:00, and if you ever question my orders again I'll have your arse sacked."
Technician: "Those are magic words. Here's your lab results."
Just an observation: if you trust that the existing wear leveling algorithm has been performing properly, you can safely reset its state as well. Think about it. Resetting them won't unbalance it by more than one extra write or so for the affected blocks.
Flash bits typically don't fail from reading, they fail from the stress of writing. And they don't wear out at exactly the 100,000th write, so you never know precisely which bit is going to be "more susceptible" to failure than another. So as long as the wear leveling algorithms work every time you write to the chip, resetting the state occasionally isn't going to impact the lifetime of the chip by significantly overwriting one block more than others. Resetting the state on a minute-by-minute basis, however, is likely to overwork the state bits first, and possibly weaken the block at the algorithm's "first block allocated from a reset state".
Your only risk is overwriting the leveling state would also reset whatever counters might signal the controller that "this is my 150,000th write, my warranty is about to expire and the chances of failure have now risen to over .000002% per write operation." I know S.M.A.R.T. monitors work with those kinds of statistics for spinning drives, but I've never bothered to look at if they report on flash drive statistics.
I don't think extra controls would have make a whit of difference in the validity of this particular study. There's simply too much variance in the environment to draw a conclusion. Valid results would be drowned in the noise of other potential natural causes of CCD: toxins, pesticides, fungi, viral infections, feral bees, predators, parasites, etc.
A proper experiment would require hundreds if not thousands of hives. And yes, they'd need better protocols for handling the phones (I like your theory about outgassing), although the idea of having the phones actually inside the hives is still absurd as a potential cause of CCD, as this experiment likely provided the world with the only two hives ever to actually have cell phones inside them.
Nope. All the bees in a single field could be impacted by a nearby external uncontrolled stimulus: anything from an early frost, drought, pesticide applications, diseases, mites, fungi, competing colonies of nearby bees, or even dirt on the shoes of the guy changing the cell phone batteries. You'd need a really large set of samples to figure out if there was even a measurable impact by using a Faraday cage. I seriously doubt that a hundred fields would provide enough samples.
Four samples would yield nothing more significant that the current article, which is another way of saying "worthless".
Should I get off your lawn now, grandpa?
Only after you pollinate some of his flowers.
The subduction zone is off the coast. How would an earthquake there affect Portland, Oregon, which is 80 miles inland?
Create a lot of new beachfront property?
Are you kidding? 99 degree weather with 99% humidity is a freakin' disaster in my book.
Yet another reason ... why I'm glad I don't live in the USA.
Fixed that for the GP. I mean, seriously. Is there a form of natural disaster you guys aren't under constant threat of?
Hey, reminding us of that constantly is the job of politicians facing elections, not slashdot.
the iPhone threw out the gimped phone device business model crammed down our throats by the phone companies.
"gimped phone device"? Holy shit, the iPhone is the most gimped phone I've had in 10 years. The lock-in to the app store? A Bluetooth stack that doesn't support OBEX or AVRCP? No J2ME?
People actually celebrated when the black turtleneckers deigned to bestow MMS upon their undeserving countenance. That's how bad Apple is, and yet you've bought the propaganda hook, line and sinker. Wake up.
I wasn't even sliding into the realm of actual criminal corporate behavior. Just working within the existing system, every activity and idea is turned into dollars before being discussed. I don't know that you can change that at any meaningful level, as that is how success is measured in the financial world. Everything is dollars (well, until the saying becomes "everything is yuan", anyway.)
I'm saying that our rights are undervalued in the current system, and the only way to change it is for us to assign them a lot more value through the act of lawsuits. If we don't sue then our rights are actually being fairly valued within the system, and they will continue to be shat upon by people like Tim O'Reilly as there is no financial penalty for trying.
Why not artificial meat in the garage? It's pretty much cloning tissue, which is possible at small lab scales. The hard part is figuring out what to clone, and how to grow it. That means lots of experiments in environmental conditions, cultural media, etc. Plenty of room for amateurs to try things, from inventing cheap yet highly stable heating systems to efficient nutrient sources.
You've hit on kind of the sweet spot there. I agree with you that the scientific world seems to have had many of its boundaries pushed beyond the capacity of the average home experimenter, but the artistic world has no such boundaries. Fun and artistic electromechanical toys and hacks are still novel. Look at shows like Burning Man, sites like hackaday, magazines like Make:. They're filled with people interested in the act of creation. And last night my brother-in-law introduced me to Farm Show magazine (farmshow.com) which is a compendium of hacks and homebuilt machines that farmers have created out of necessity and imagination. It has a lot of really cool homemade things in it.
And if you're looking to monetize it, handmade and homemade mechanical equipment has a very visceral appeal to a lot of people. The potential to sell a unique device is high. And you can get involved for any amount of money, from repurposing junk bits from broken VCRs to building a nicely equipped machine shop.
My point is that corporations will continue to assess personal privacy issues as low-dollar-risk line items, even if a few executives go to jail. There's personal incentive for one or two people, but not the corporation as a whole. "Let's continue to screw with people's privacy but let's set up a scapegoat to take the fall" then becomes the unwritten corporate strategy. A jailed CTO won't result in a multimillion dollar hit to the shareholders, as much as the CTO's ego wishes it to be true.
But if analysts and investors are convinced that messing with privacy issues makes stock worthless, they will stop funding companies who try to violate them. Every stinking decision on the Street is based on the almighty dollar.
Small companies would retain the equal opportunity to be sued into oblivion. They'd just be forgotten faster than an operation the size of Facebook.
Just one effective lawsuit against a mom'n'pop shop would likely result in the personal bankruptcy of the owners. That's never a pleasant thought to a small business owner, which is why I'd trust them to have better intentions to do right by their customers. Of course as a mom'n'pop, I'd also not be surprised by a weaker implementation of security measures.
Sorry, but that won't work. As a typical amoral shareholder or mutual funds investor, I'm perfectly OK with the CEO going to jail as long as my dividend checks keep arriving in the mailbox. The market runs only on greed, not fear of incarceration of "other people".
But imagine what would happen if Facebook was sued out of existence because of this change to their privacy policy. The next company to talk about loosening their privacy policy would see their share value dropping in half, as the wary investors divest as fast as they can.
What needs to happen is that a lot of people who were adversely impacted by this have to file giant lawsuits. Let's say that ten thousand people lost their jobs by having "drunk college pictures" revealed to their employers. If I were one of them, I'd sue for lost wages on a career where I would have potentially made $5,000,000 over my lifetime. Get a thousand victims to file those lawsuits, and the company collapses under the burden. Message to corporations: if you mess with our privacy, we will parade your rotting corpse down Wall Street and toss it in the Hudson River.
This is still America. We have the right to file lawsuits if we're aggrieved. It's the only mechanism by which individuals can teach corporations lessons that will stick.
It's become the fashion to lump everything together, as if performances, images, tangible goods, rights, efforts, and ideas are all exactly the same kinds of assets and should be treated exactly the same by corporations, governments, and individuals. That's happened because business students are taught to convert everything to dollars, assign a value to risk, and then simply slide the numbers around on an Excel spreadsheet until the biggest one pops out at the bottom.
The problem is that the dollar value they assign to risk is based on the imaginations of some not-very-creative people, and is only the risk to them, not to the end users. "Well, if we screw with the privacy policy, our risk is that we'll lose less than 0.5% of our users. That's equal to ad revenue of $3,000,000. The ROI on increased ads is projected to be $10,000,000. This change will pay itself back in months, so just do it."
What really has to happen is truckloads and truckloads of lawsuits have to be filed against them, by people whose privacy was violated. They have to learn that if they mess with our privacy, it will cost them billions of dollars in settlements and legal fees -- to the brink of bankruptcy, and even over the edge. That is the only time corporations will start respecting our rights -- when violating them is guaranteed to flush their bottom lines into the toilet.
My 20-year-old rot13 script didn't survive the transition from HP/UX to Linux. I put the cause down to
wait for it...
...
bit-rot.
Thank you, I'll be here all the week. Try the fish.