The RFID tag is, at the end of the day, a physical artifact. It's a very small one, and very cheap one, but that's what it is. It's an item, made in a factory, which is attached to other things in order to identify them. Like... well, a tag.
But bar codes aren't tags. They're not artifact-based. They're just information. Encoded information. They get printed on a physical artifact in most cases, but that's not the point. No physical object is required other than the object you want to identify, assuming you can print onto it.
The cost of adding a bar code to an envelope is essentially nil, once you have the system set up. To put an RFID tag on, not only do you need to set up the system, but you are dependent on a constant supply of physical artifacts: the tags themselves. For this reason alone, we're going to see bar codes hanging around for a long, long time.
You need to calm down. Maybe take a few of those pills you were prescribed.
RFID tags don't allow remote tracking and location, any more than I can tell where your car is right now because you have a EasyPass toll card on the windshield, or a Mobil SpeedPass on your keychain. Sure, if I had access to the requisite databases perhaps I could find out what interstate you last travelled on, or what gas station you last filled up at, but there's no way to tell where an RFID tag is when it's not near a scanner.
I'm not sure where people get the idea that RFID tags are like little homing beacons, somehow reporting their position back to Ground Control constantly. As neat as that would be (and people are actively working on it, check out http://www.woz.com/ it's not here yet. And RFID certainly isn't it.
It's unfortunate Slashdot wasn't around when barcodes were introduced. I'd like to be able to link back to posts where people would have claimed that barcodes are going to allow the Man to track your every move, find out if you're a Red sympathizer, reveal if you're gay.
Whenever a new technology comes out, first legitimate issues are raised, and then the technology is exaggerated beyond all recognition, as are the concerns. And the sad part about it all is that the original -- and valid -- privacy concerns are almost completely overrun by half-baked claims.
You do realize that most RFID tags derive their power from the RF field put out by the scanner, right?
You need to have a fairly powerful field to get them to transmit at all. I can't imagine how much power you'd need to put into a UHF tag in order to get it to send back a signal that you could read 20' away, even in free space. You've got the inverse square law working against you both ways.
I think you'd probably tip your neighbor off when you blew out their TV, microwave, computer, toaster, and everything else with a chip in it, with the huge amount of EMF you'd need to hit all the UHF/microwave RFID tags on his groceries.
I know that this whole conspiracy thing is fun to think about, but please people, let's be practical. In all seriousness, I'm sure that there is a very hard upper limit on the range of these tags, beyond which you'd melt the chip in order to get enough power out of it to transmit a signal of the required strength back to you.
It also helps the manufacturers, but in an indirect way: it prevents mis-shipments.
What a lot of people don't realize is that if you're a company who sells to Wal-Mart, and you send a shipment of product to the wrong store, you're really hosed. First of all, you have to immediately reship the product to the correct store. Secondly, the product that you accidentally shipped doesn't come back to you. The store that got it keeps it, and just tries to liquidate it as quickly as possible.
So if you're a manufacturer and want to do business with Wal-Mart, and despite their reputation for being tough to work with, a lot of manufacturers do, anything that cuts down on mis-shipments is a really good thing.
As I said in another post, the technology to do item-by-item RFID tagging isn't here yet. There are too many technological problems yet to solve, or rather there are economic problems justifying the outlay of cash that solving these technological problems would take.
Grocery and retail stores might take a brief look at RFID as a way to prevent shoplifting, but where they're really interested in a solution (and where they really lose money) is in back room inventory 'shrink.' This is where you have a room full of DVD players on Monday, and when you finally go to move them out on the floor at the end of the week, you're short a dozen units for "no apparent reason."
What they want to do is put RFID tags on the cases, if they're big enough, or on the pallets, if not. Then you put a scanner at the loading dock at the distribution center or warehouse, to scan the items as they're put on the truck, another at the loading dock at the store, to count them as they come OFF the truck, and another at the stockroom door to count them as they go onto the sales floor. It can't stop theft/shrink completely, but it can at least tell you where it's happening. Are they getting stolen out of the distribution center before they're on the truck? Out of the truck in transit? Or from the back room by employees?
Shoplifting is a problem, but it's not nearly as big as the behind-the-scenes theft that can occur, especially when you've got trucks full of high value density merchandise that's not well controlled, and where it can take a long time and a long ways down the supply line to find out that units are missing.
I'll offer a counterexample. A few years ago I got curious as to whether WalMart, which is where everyone from my uni at the time bought EVERYTHING, was cheaper than the K-Mart located right next door.
I really wanted the K-Mart to win, too, for personal reasons. Mostly because the Wal-Mart had just expanded into a Super Wal-Mart (dare I say supersized?) and left an empty building across the street. This irked the hell out of me. So I was set on proving that there was no reason to shop there.
I put together a list that I thought represented about a shopping cart's worth of stuff that an average university student would buy. So not necessarily household items per se, but some food items, consumables like tissues, paper towels, and plastic cups, sheets and towels, few pieces of cheap furniture, etc.
Anyway, to make a long story short, despite my desire to stick one to WM, items at K-Mart really were more expensive. There wasn't any clear pattern to the price difference, but in general WM was the same price or less than K-Mart, sometimes by a few dollars. Note we weren't using name-brand stuff, we went for whatever we deemed to be about comparable generic equivalents. It's certainly not a scientific study, but it really shot my whole theory about Wal-Mart being a scam out from under me.
I suppose it's possible that Wal-Mart was purposely undercutting the K-Mart in order to drum up customers since it had just supersized itself, and perhaps these discrepancies averaged out in time, but I doubt it. It's been a few years now, I'd be interested to see the same thing done over again.
In the end, you have to ask yourself how it was that Wal-Mart has managed to displace so many of the traditional big box stores. I'm sure some of it was due to aggressive business tactics, but I think it's mostly due to the fact that they just have a supply chain that's orders of magnitude better than any of their competitors were, back when they entered the market. This meant they could undercut them just enough to win consumers, and capture marketshare. As much as I dislike a lot of things about them, they didn't get huge using some sort of redneck mind-control ray, which is what a lot of anti-WM people seem to believe.
I used to work at a very small grocery store where there wasn't a bar code scanner. There might have been one built in to the cash register, buried under the accumulated detritus that surrounded the register, but I never saw it working. Instead the cashiers just memorized the prices for almost the items in the store.
When I first started working there and saw the book of prices I was expected to memorize, I was incredulous. But after a few days I found that I rarely needed to look in the book, except for very obscure items. After a few months, I forgot where the book was.
It's actually pretty amazing how much information you can store, when you're learning it as you're working instead of just sitting down and trying to memorize a list.
OT: I actually went back to that store just recently, and noticed that they had tossed the old register and book of prices, and replaced it with a very fancy barcode-scanner/scale/touchscreen POS console with integrated credit card acceptor, check and receipt printer, etc. Actually very excessive for such a little grocery store. But the kid working what would have been my job at the register was still bitching at it constantly, as he stood there and just did nothing but flick the groceries by the scanner on their way to the bag.
Has anybody actually bothered to look at the frequencies that those Wal-Mart RFID tags use?
I'm almost certain they're in the UHF and up range. I wouldn't be surprised if they're in the 2.4GHz ISM band either (like everything else these days). But for anyone that's never tried using their wifi AP in a basement or metal framed building, those waves don't much like metal or very solid objects. And RFID transmitters aren't very strong either.
I can see item-by-item tagging working if you only have an entire case of something like toilet paper or paper towels, or maybe even clothes... but imagine trying to read the tag that's on an item buried in the center of a cart full of Campbells Condensed Tomato Soup? You'd be boiling the soup in the outer cans before you got enough power to the tags on the inner items to energize them.
I think item-by-item tagging is a ways off. There's a reason they're only putting these things on cardboard boxes right now.
Currently, the speed of the belts at a place like WalMart or FedEx, or any other facility heavily dependent on bar codes, is how fast employees can flip the boxes.
As they come past on the belt, you have a bunch of people standing around, who orient the boxes so that their bar codes are facing one of the directions the scanner will read (I think they read on three sides, but the more complicated 2-D barcodes may only be able to be read from one direction, I'm not sure). At top speed on one of those conveyors, you need to have several people standing in a line by the belt to flip the boxes as they come past. If you have four people, each person only needs to flip every fourth box.
The advantage to RFID is no flipping -- it doesn't matter what orientation the boxes go through the scanner at, so you can take those four 'flippers' and have them unload the trucks faster. You can pretty much have a solid line of boxes going into the scanner.
Perhaps what I should have said was "I doubt they'd tolerate a system that scanned all their archived email in order to produce publicly available graphical model, for academic or curiosity purposes.":)
I don't pretend to have any insight into what the people at those agencies do, but I agree that if I worked there, I'd definitely have approached Google by now. Having that much important information in one place, indexed and searchable, is just too much of an opportunity for an entity that essentially lives by sifting nuggets of data from seas of irrelevant noise to pass up.
That's very true. In all this discussion about GMail it sort of slipped my mind that there's no reason you couldn't do the same thing with any other mail reader.
I guess the reason for GMail in this situation was because it retains all email by default, while I think many people who use standard offline readers delete most of their mail, while only retaining important bits. That would really skew the output of the model (although admittedly it might skew it in a very interesting direction).
Writing such a program would be a bit out of my league, but I'd also be interested to run it on my own archive just to see the output. Of course, running it on just your own mail wouldn't produce terribly interesting a "network," since there would just be one central node and then a lot of 1st degree nodes radiating away from that, at various angles and distances. To produce a network you'd need to be able to search recursively down the node structure, which as other people pointed out, raises a lot of privacy and security concerns.
You're wrong to blame Google for it, though. All Google is doing is making the technology of surveillance more obvious. Your emails and IMs aren't, and never were, private. Unless you were using some form of end-to-end encryption, that is. But for the vast majority of people, that assumption of privacy, at least when it comes to the Internet, is just that: an assumption. And a very poor one at that.
Frankly, I like GMail. I think everyone ought to use it. Okay, not really. But I like that it makes people like my parents, who despite years of cautioning never gave a second thought about emailing someone their bank routing number or Amex-online account login, think twice about what they type in. You can rail all day to people about how email is really nothing more secure than a postcard, passed from machine to machine across the network, but that's all very abstract. The first time you notice how those GMail ads seem to eerily change depending upon what you're writing about, the whole thing becomes more clear.
Google isn't invading your privacy. It's just making you aware of the fact that you never had any.
Of course, people say, before Google existed and thousands of users' emails were archived and indexed, intercepting email was hard. Okay, point granted. But really how hard? Certainly not outside the reach of government agencies. If you're really afraid of the three-letter-guys, then everything Google does to drive the unencrypted=insecure link home to the average user is good.
Because the only privacy you'll get on the internet is the kind you create for yourself. The more users who realize this and the sooner they take steps to implement it, the better. When everyone starts actually encrypting their email and messaging, then we'll actually have some privacy for the government to try and invade.
I've played around with a "social networking" site (thefacebook.com) which used to have -- and perhaps it's back now, I'm not sure -- a really neat feature. It would give you a zoomable 2-D representation of your social network as an SVG file.
The SVG part isn't important (I was never really clear why they picked this format over any other vector format...) but the idea was pretty cool. With yourself at the center you could see who was near you in the social structure of the network, and who they were closely associated with, etc.
I never used the service long or heavily enough to develop a very robust social network (and not enough of my friends used it to make a very good model of the real world) but this is not so with email. Practically everyone I know, with the exception of a few older or particularly Luddite relatives, I've emailed at one point or another.
It would be neat to have a program that scanned your email archives (with Gmail this wouldn't be too hard, since all mail is retained in the archive unless manually deleted) and constructed a social network from it. If I were going to design it, I'd make "closeness" be the frequency of emails, and the angular separation between two people who both talk to a third based on the number of shared keywords in their emails. That way you'd end up with all your business associates off in one direction (say all radiating away from you within a few degrees of each other) but your family, with whom you probably use few words in common with your business emails, in a different direction entirely.
If the program could scan people's emails recursively -- assuming they were all on Gmail and had suitably large archives of email -- you could create a pretty neat social model that would actually be reflective of the real world.
Of course, the privacy issues surrounding something like that would be gigantic. People get all creeped out by Gmail scanning emails and then presented targeted advertising...I doubt they'd tolerate a system that scanned all their archived email in order to produce a graphical model, even if it was semi-anonymous.
If you really live an absolutely virtuous, by-the-books, 100% wholesome American life, then by all means you can shout at me about how I use HandBrake to rip DVDs to MP4s so I can watch them without filling my entire hard drive.
But if you've recently exceed the speed limit, littered, drove in the HOV lane with less than two passengers, or any other multitude of other things that I see people doing every day which are a hell of a lot worse, in terms of direct effects, than my DVD ripping, shut the hell up.
There's a quote about people living in glass houses which is appropriate to this situation. If you don't know which one I'm talking about, ask your mother.
Everything you say is true, except for one bit, when you say the words "re-rip lossless." While you can definitely re-rip the burned CD in a lossless format, the entire burn-out and rip-in transaction WILL incur a serious sonic loss. If all you're ever going to do is play the music on iPod headphones, a budget car stereo, or computer speakers, you might never notice though, but it's there.
This might be completely untrue, but I was told once it had something to do with filename length restrictions in the iPod's embedded OS.
The song names that you see aren't taken from the file names, they're taken from the ID3 tags, and from a database which cross references song names to file names.
Thus, the iPod's embedded system never has to deal with long file names, which are pretty common if you name your music according to the "[Artist] - [Song].mp3" form, especially if you don't abbreviate anything.
This might be completely wrong, but it's the best explanation I've ever heard of that particular oddity. The iPod can carry files with long names just fine, but the internal software doesn't ever work with them.
This might be an extremely ignorant question, but how can that work?
Perhaps I misunderstand, but I thought the point of the website's hash table was so that when a user wants to log in later on, you hash the new entry attempt, and compare it to the stored hash. If they match, access granted.
If you included some random salt in the original hash, how would you ever compare it to an entered password attempt later down the line, without also storing a list of the salts?
And if you store a list of the salts, how are you creating any additional security? The only additional piece of information you're adding to the hash cannot be secret...so what is its purpose?
The point here isn't the hardware they ran it on, but rather the software they designed to make a workstation that's truly hardware (or at least processor-and-mobo) independent. You can take the box away, plug it into a different workstation somewhere, and it comes right back up as if it was your own computer.
Think about how many employees IBM has worldwide. It's in the hundreds of thousands. Think of how much it costs to equip them all with Thinkpads (even if they are made by Lenovo now). Now think about the money you could save if even a small percentage of those people could get a $200 box instead of a laptop, and just plug it into any available desktop. That's the think driving development here, I think.
That they chose the iPod to demo the software on probably has a lot more to do with showmanship and use of budget than any technical requirement. Any FW drive would have worked... buy why get a boring old FW drive when you can get a cool sexy white one that every executive will remember and know about, and also is cool to use to play music on when you're done with the project?
As others have pointed out, in the U.S. what matters is who invented the device or process first, not who filed the patent. The patent of course is easy proof that you had invented the device by the filing date, but it doesn't preclude someone from challenging the patent by showing evidence that they invented it first. I believe that the burden of proof is on the challenger of the patent, though, to show they invented it earlier.
Also as others have pointed out, the "earlier" invention needs to encompass all of the elements patented, not just a few, in order to invalidate it. Whether MS was smart enough to think of that and include in their patent some things which either aren't used on the iPod or weren't added until recently, I don't know.
Although I think Apple is clearly in the right here, and the legal system will be on their side, the whole situation concerns me. It just seems like one of those things that shouldn't be able to happen.
Funny how those things so often involve Microsoft, eh?
No, the way you deal with most government documents, in my experience, is that you SIGN the original, preferably in blue ink, and then photocopy it five times for the "copies."
There used to be places where all documents had to be signed in blue ink, to distinguish them from photocopies, but I think those rules have either depreciated or are no longer enforced.
A lot of times with military orders that I used to see, an original hardcopy was signed and kept on file somewhere, and then the electronic or faxed-copy went out with "[ORIGINAL SIGNED]" in the signature field of the document. Not particularly secure, but then again these weren't anything interesting.
They would have thought about it, for about fifteen seconds, and then realized that if they sold a single computer with Apple's OS, that Microsoft would pull the steep discount they get on Windows licenses (and which allows them to undercut the competition and maintain their marketshare).
If Apple did release an operating system for commodity hardware, it would be the Power Computings and UMaxes of the world that would be selling the systems. MS would never let the major manufacturers touch it.
Slashdot Mirror
This is a good point.
... well, a tag.
The RFID tag is, at the end of the day, a physical artifact. It's a very small one, and very cheap one, but that's what it is. It's an item, made in a factory, which is attached to other things in order to identify them. Like
But bar codes aren't tags. They're not artifact-based. They're just information. Encoded information. They get printed on a physical artifact in most cases, but that's not the point. No physical object is required other than the object you want to identify, assuming you can print onto it.
The cost of adding a bar code to an envelope is essentially nil, once you have the system set up. To put an RFID tag on, not only do you need to set up the system, but you are dependent on a constant supply of physical artifacts: the tags themselves. For this reason alone, we're going to see bar codes hanging around for a long, long time.
You need to calm down. Maybe take a few of those pills you were prescribed.
RFID tags don't allow remote tracking and location, any more than I can tell where your car is right now because you have a EasyPass toll card on the windshield, or a Mobil SpeedPass on your keychain. Sure, if I had access to the requisite databases perhaps I could find out what interstate you last travelled on, or what gas station you last filled up at, but there's no way to tell where an RFID tag is when it's not near a scanner.
I'm not sure where people get the idea that RFID tags are like little homing beacons, somehow reporting their position back to Ground Control constantly. As neat as that would be (and people are actively working on it, check out http://www.woz.com/ it's not here yet. And RFID certainly isn't it.
It's unfortunate Slashdot wasn't around when barcodes were introduced. I'd like to be able to link back to posts where people would have claimed that barcodes are going to allow the Man to track your every move, find out if you're a Red sympathizer, reveal if you're gay.
Whenever a new technology comes out, first legitimate issues are raised, and then the technology is exaggerated beyond all recognition, as are the concerns. And the sad part about it all is that the original -- and valid -- privacy concerns are almost completely overrun by half-baked claims.
You do realize that most RFID tags derive their power from the RF field put out by the scanner, right?
You need to have a fairly powerful field to get them to transmit at all. I can't imagine how much power you'd need to put into a UHF tag in order to get it to send back a signal that you could read 20' away, even in free space. You've got the inverse square law working against you both ways.
I think you'd probably tip your neighbor off when you blew out their TV, microwave, computer, toaster, and everything else with a chip in it, with the huge amount of EMF you'd need to hit all the UHF/microwave RFID tags on his groceries.
I know that this whole conspiracy thing is fun to think about, but please people, let's be practical. In all seriousness, I'm sure that there is a very hard upper limit on the range of these tags, beyond which you'd melt the chip in order to get enough power out of it to transmit a signal of the required strength back to you.
It also helps the manufacturers, but in an indirect way: it prevents mis-shipments.
What a lot of people don't realize is that if you're a company who sells to Wal-Mart, and you send a shipment of product to the wrong store, you're really hosed. First of all, you have to immediately reship the product to the correct store. Secondly, the product that you accidentally shipped doesn't come back to you. The store that got it keeps it, and just tries to liquidate it as quickly as possible.
So if you're a manufacturer and want to do business with Wal-Mart, and despite their reputation for being tough to work with, a lot of manufacturers do, anything that cuts down on mis-shipments is a really good thing.
And you wonder why his country got overrun by Mongols....?
As I said in another post, the technology to do item-by-item RFID tagging isn't here yet. There are too many technological problems yet to solve, or rather there are economic problems justifying the outlay of cash that solving these technological problems would take.
Grocery and retail stores might take a brief look at RFID as a way to prevent shoplifting, but where they're really interested in a solution (and where they really lose money) is in back room inventory 'shrink.' This is where you have a room full of DVD players on Monday, and when you finally go to move them out on the floor at the end of the week, you're short a dozen units for "no apparent reason."
What they want to do is put RFID tags on the cases, if they're big enough, or on the pallets, if not. Then you put a scanner at the loading dock at the distribution center or warehouse, to scan the items as they're put on the truck, another at the loading dock at the store, to count them as they come OFF the truck, and another at the stockroom door to count them as they go onto the sales floor. It can't stop theft/shrink completely, but it can at least tell you where it's happening. Are they getting stolen out of the distribution center before they're on the truck? Out of the truck in transit? Or from the back room by employees?
Shoplifting is a problem, but it's not nearly as big as the behind-the-scenes theft that can occur, especially when you've got trucks full of high value density merchandise that's not well controlled, and where it can take a long time and a long ways down the supply line to find out that units are missing.
Care to justify this "proven fact?"
I'll offer a counterexample. A few years ago I got curious as to whether WalMart, which is where everyone from my uni at the time bought EVERYTHING, was cheaper than the K-Mart located right next door.
I really wanted the K-Mart to win, too, for personal reasons. Mostly because the Wal-Mart had just expanded into a Super Wal-Mart (dare I say supersized?) and left an empty building across the street. This irked the hell out of me. So I was set on proving that there was no reason to shop there.
I put together a list that I thought represented about a shopping cart's worth of stuff that an average university student would buy. So not necessarily household items per se, but some food items, consumables like tissues, paper towels, and plastic cups, sheets and towels, few pieces of cheap furniture, etc.
Anyway, to make a long story short, despite my desire to stick one to WM, items at K-Mart really were more expensive. There wasn't any clear pattern to the price difference, but in general WM was the same price or less than K-Mart, sometimes by a few dollars. Note we weren't using name-brand stuff, we went for whatever we deemed to be about comparable generic equivalents. It's certainly not a scientific study, but it really shot my whole theory about Wal-Mart being a scam out from under me.
I suppose it's possible that Wal-Mart was purposely undercutting the K-Mart in order to drum up customers since it had just supersized itself, and perhaps these discrepancies averaged out in time, but I doubt it. It's been a few years now, I'd be interested to see the same thing done over again.
In the end, you have to ask yourself how it was that Wal-Mart has managed to displace so many of the traditional big box stores. I'm sure some of it was due to aggressive business tactics, but I think it's mostly due to the fact that they just have a supply chain that's orders of magnitude better than any of their competitors were, back when they entered the market. This meant they could undercut them just enough to win consumers, and capture marketshare. As much as I dislike a lot of things about them, they didn't get huge using some sort of redneck mind-control ray, which is what a lot of anti-WM people seem to believe.
I used to work at a very small grocery store where there wasn't a bar code scanner. There might have been one built in to the cash register, buried under the accumulated detritus that surrounded the register, but I never saw it working. Instead the cashiers just memorized the prices for almost the items in the store.
When I first started working there and saw the book of prices I was expected to memorize, I was incredulous. But after a few days I found that I rarely needed to look in the book, except for very obscure items. After a few months, I forgot where the book was.
It's actually pretty amazing how much information you can store, when you're learning it as you're working instead of just sitting down and trying to memorize a list.
OT: I actually went back to that store just recently, and noticed that they had tossed the old register and book of prices, and replaced it with a very fancy barcode-scanner/scale/touchscreen POS console with integrated credit card acceptor, check and receipt printer, etc. Actually very excessive for such a little grocery store. But the kid working what would have been my job at the register was still bitching at it constantly, as he stood there and just did nothing but flick the groceries by the scanner on their way to the bag.
Has anybody actually bothered to look at the frequencies that those Wal-Mart RFID tags use?
... but imagine trying to read the tag that's on an item buried in the center of a cart full of Campbells Condensed Tomato Soup? You'd be boiling the soup in the outer cans before you got enough power to the tags on the inner items to energize them.
I'm almost certain they're in the UHF and up range. I wouldn't be surprised if they're in the 2.4GHz ISM band either (like everything else these days). But for anyone that's never tried using their wifi AP in a basement or metal framed building, those waves don't much like metal or very solid objects. And RFID transmitters aren't very strong either.
I can see item-by-item tagging working if you only have an entire case of something like toilet paper or paper towels, or maybe even clothes
I think item-by-item tagging is a ways off. There's a reason they're only putting these things on cardboard boxes right now.
Currently, the speed of the belts at a place like WalMart or FedEx, or any other facility heavily dependent on bar codes, is how fast employees can flip the boxes.
As they come past on the belt, you have a bunch of people standing around, who orient the boxes so that their bar codes are facing one of the directions the scanner will read (I think they read on three sides, but the more complicated 2-D barcodes may only be able to be read from one direction, I'm not sure). At top speed on one of those conveyors, you need to have several people standing in a line by the belt to flip the boxes as they come past. If you have four people, each person only needs to flip every fourth box.
The advantage to RFID is no flipping -- it doesn't matter what orientation the boxes go through the scanner at, so you can take those four 'flippers' and have them unload the trucks faster. You can pretty much have a solid line of boxes going into the scanner.
Here's all you need to know from that article, or that anyone else needs to know from your post:
"mute point" = WRONG.
That's not a bug. It's a feature.
You make a good point.
:)
Perhaps what I should have said was "I doubt they'd tolerate a system that scanned all their archived email in order to produce publicly available graphical model, for academic or curiosity purposes."
I don't pretend to have any insight into what the people at those agencies do, but I agree that if I worked there, I'd definitely have approached Google by now. Having that much important information in one place, indexed and searchable, is just too much of an opportunity for an entity that essentially lives by sifting nuggets of data from seas of irrelevant noise to pass up.
That's very true. In all this discussion about GMail it sort of slipped my mind that there's no reason you couldn't do the same thing with any other mail reader.
I guess the reason for GMail in this situation was because it retains all email by default, while I think many people who use standard offline readers delete most of their mail, while only retaining important bits. That would really skew the output of the model (although admittedly it might skew it in a very interesting direction).
Writing such a program would be a bit out of my league, but I'd also be interested to run it on my own archive just to see the output. Of course, running it on just your own mail wouldn't produce terribly interesting a "network," since there would just be one central node and then a lot of 1st degree nodes radiating away from that, at various angles and distances. To produce a network you'd need to be able to search recursively down the node structure, which as other people pointed out, raises a lot of privacy and security concerns.
You're right to be concerned.
You're wrong to blame Google for it, though. All Google is doing is making the technology of surveillance more obvious. Your emails and IMs aren't, and never were, private. Unless you were using some form of end-to-end encryption, that is. But for the vast majority of people, that assumption of privacy, at least when it comes to the Internet, is just that: an assumption. And a very poor one at that.
Frankly, I like GMail. I think everyone ought to use it. Okay, not really. But I like that it makes people like my parents, who despite years of cautioning never gave a second thought about emailing someone their bank routing number or Amex-online account login, think twice about what they type in. You can rail all day to people about how email is really nothing more secure than a postcard, passed from machine to machine across the network, but that's all very abstract. The first time you notice how those GMail ads seem to eerily change depending upon what you're writing about, the whole thing becomes more clear.
Google isn't invading your privacy. It's just making you aware of the fact that you never had any.
Of course, people say, before Google existed and thousands of users' emails were archived and indexed, intercepting email was hard. Okay, point granted. But really how hard? Certainly not outside the reach of government agencies. If you're really afraid of the three-letter-guys, then everything Google does to drive the unencrypted=insecure link home to the average user is good.
Because the only privacy you'll get on the internet is the kind you create for yourself. The more users who realize this and the sooner they take steps to implement it, the better. When everyone starts actually encrypting their email and messaging, then we'll actually have some privacy for the government to try and invade.
I've played around with a "social networking" site (thefacebook.com) which used to have -- and perhaps it's back now, I'm not sure -- a really neat feature. It would give you a zoomable 2-D representation of your social network as an SVG file.
The SVG part isn't important (I was never really clear why they picked this format over any other vector format...) but the idea was pretty cool. With yourself at the center you could see who was near you in the social structure of the network, and who they were closely associated with, etc.
I never used the service long or heavily enough to develop a very robust social network (and not enough of my friends used it to make a very good model of the real world) but this is not so with email. Practically everyone I know, with the exception of a few older or particularly Luddite relatives, I've emailed at one point or another.
It would be neat to have a program that scanned your email archives (with Gmail this wouldn't be too hard, since all mail is retained in the archive unless manually deleted) and constructed a social network from it. If I were going to design it, I'd make "closeness" be the frequency of emails, and the angular separation between two people who both talk to a third based on the number of shared keywords in their emails. That way you'd end up with all your business associates off in one direction (say all radiating away from you within a few degrees of each other) but your family, with whom you probably use few words in common with your business emails, in a different direction entirely.
If the program could scan people's emails recursively -- assuming they were all on Gmail and had suitably large archives of email -- you could create a pretty neat social model that would actually be reflective of the real world.
Of course, the privacy issues surrounding something like that would be gigantic. People get all creeped out by Gmail scanning emails and then presented targeted advertising...I doubt they'd tolerate a system that scanned all their archived email in order to produce a graphical model, even if it was semi-anonymous.
This point can't be made often enough I think.
If you really live an absolutely virtuous, by-the-books, 100% wholesome American life, then by all means you can shout at me about how I use HandBrake to rip DVDs to MP4s so I can watch them without filling my entire hard drive.
But if you've recently exceed the speed limit, littered, drove in the HOV lane with less than two passengers, or any other multitude of other things that I see people doing every day which are a hell of a lot worse, in terms of direct effects, than my DVD ripping, shut the hell up.
There's a quote about people living in glass houses which is appropriate to this situation. If you don't know which one I'm talking about, ask your mother.
Plutonium and solid rocket fuel.
It'll be a very short war. About 25 minutes, I think.
Everything you say is true, except for one bit, when you say the words "re-rip lossless." While you can definitely re-rip the burned CD in a lossless format, the entire burn-out and rip-in transaction WILL incur a serious sonic loss. If all you're ever going to do is play the music on iPod headphones, a budget car stereo, or computer speakers, you might never notice though, but it's there.
This might be completely untrue, but I was told once it had something to do with filename length restrictions in the iPod's embedded OS.
The song names that you see aren't taken from the file names, they're taken from the ID3 tags, and from a database which cross references song names to file names.
Thus, the iPod's embedded system never has to deal with long file names, which are pretty common if you name your music according to the "[Artist] - [Song].mp3" form, especially if you don't abbreviate anything.
This might be completely wrong, but it's the best explanation I've ever heard of that particular oddity. The iPod can carry files with long names just fine, but the internal software doesn't ever work with them.
This might be an extremely ignorant question, but how can that work?
Perhaps I misunderstand, but I thought the point of the website's hash table was so that when a user wants to log in later on, you hash the new entry attempt, and compare it to the stored hash. If they match, access granted.
If you included some random salt in the original hash, how would you ever compare it to an entered password attempt later down the line, without also storing a list of the salts?
And if you store a list of the salts, how are you creating any additional security? The only additional piece of information you're adding to the hash cannot be secret...so what is its purpose?
The point here isn't the hardware they ran it on, but rather the software they designed to make a workstation that's truly hardware (or at least processor-and-mobo) independent. You can take the box away, plug it into a different workstation somewhere, and it comes right back up as if it was your own computer.
... buy why get a boring old FW drive when you can get a cool sexy white one that every executive will remember and know about, and also is cool to use to play music on when you're done with the project?
Think about how many employees IBM has worldwide. It's in the hundreds of thousands. Think of how much it costs to equip them all with Thinkpads (even if they are made by Lenovo now). Now think about the money you could save if even a small percentage of those people could get a $200 box instead of a laptop, and just plug it into any available desktop. That's the think driving development here, I think.
That they chose the iPod to demo the software on probably has a lot more to do with showmanship and use of budget than any technical requirement. Any FW drive would have worked
Not true at all.
As others have pointed out, in the U.S. what matters is who invented the device or process first, not who filed the patent. The patent of course is easy proof that you had invented the device by the filing date, but it doesn't preclude someone from challenging the patent by showing evidence that they invented it first. I believe that the burden of proof is on the challenger of the patent, though, to show they invented it earlier.
Also as others have pointed out, the "earlier" invention needs to encompass all of the elements patented, not just a few, in order to invalidate it. Whether MS was smart enough to think of that and include in their patent some things which either aren't used on the iPod or weren't added until recently, I don't know.
Although I think Apple is clearly in the right here, and the legal system will be on their side, the whole situation concerns me. It just seems like one of those things that shouldn't be able to happen.
Funny how those things so often involve Microsoft, eh?
No, the way you deal with most government documents, in my experience, is that you SIGN the original, preferably in blue ink, and then photocopy it five times for the "copies."
There used to be places where all documents had to be signed in blue ink, to distinguish them from photocopies, but I think those rules have either depreciated or are no longer enforced.
A lot of times with military orders that I used to see, an original hardcopy was signed and kept on file somewhere, and then the electronic or faxed-copy went out with "[ORIGINAL SIGNED]" in the signature field of the document. Not particularly secure, but then again these weren't anything interesting.
No, Dell wouldn't be selling Apple anything.
They would have thought about it, for about fifteen seconds, and then realized that if they sold a single computer with Apple's OS, that Microsoft would pull the steep discount they get on Windows licenses (and which allows them to undercut the competition and maintain their marketshare).
If Apple did release an operating system for commodity hardware, it would be the Power Computings and UMaxes of the world that would be selling the systems. MS would never let the major manufacturers touch it.