How are they supposed to be compatible with previous drives??? IDE. SCSI. Or if you want to look forward instead of backwards, Firewire or USB 2.0. The article is lacking in technical details, but my best guess is that they just repackaged IDE with a custom connector, just so anyone foolish enough to buy one is locked in.
I still can't figure out why the LS-120 has failed to take off
1. Marketing. For example, the customer that told me we had to get a zip drive so they could mail us disks probably never heard of the LS-120. (Of course, now we just use CD-R and almost any computer can read the disks.
2. That compatibility comes at a price. A have had problems now and then the 1.44 meg floppies because the case is so thin it flexes more than it should and apparently the drive heads can't find the tracks. I've also had instances where the drive itself would flex too much when you bolted it in. By abandoning the standard form factors, Iomega could put massive cases on the disk and the drive, so they will never flex. The LS-120 is as thin and not-quite-rigid as the 1.44's, so unless they have compensated for this somehow, I wouldn't expect high reliability.
3. By now, the time for super-floppies has come and gone. The drives are only a little cheaper than CD-R/RW drives, the disks are a lot more expensive, and with CD-R I can create a disk that you don't need a special drive to read.
When cordless phones first became common, many people were surprised to discover that their neighbors were listening in. DUH!!!
When cellular phones came within ordinary peoples' price range, many were surprise to learn that everyone could listen in. DUH!!!
Anything you put on the radio is insecure unless it is heavily encrypted with good control of the keys. Why is that hard to understand?
The wireless keyboard and mouse could be encrypted. In fact from the article it appears that they might be encrypted; there is some sort of negotiation going on at startup, but I don't know whether that is to pick a key or simply to pick a channel. But even if the encryption is good, this live on-the-air key negotiation is a weak point. For instance, you could buy the same model of keyboard and take control after the guy turned on his computer and while he was walking over to the keyboard. Of course, you'd be entering commands blind, but there's always "del *.* (enter) y". Or since there seems to be a short list of built-in keys, you could experiment with a keyboard to find out what they all were, read the key selected from the start-up transmissions, then read out the login and password.
If you want a really secure wireless connection, then you need strong encryption with a unique key that no one else knows. Either you ship keyboard and receiver from the factory as a set (and trust the factory to erase the pre-programmed keys from their records as soon as they are used), or you have a way to temporarily bring the two devices together and connect them by a nearly untappable wire while they figure out a key.
Finally, there is a mathematical procedure that is claimed to work out a secure key by a _long_ process of exchanged messages and intensive calculation. Don't ask me to explain it. It would require enabling two-way communications, which doubles the cost of the radio circuits, and I suspect it would increase the CPU power required dramatically.
By the way, you don't need much CPU power for good secret-key encryption, you just have to design right. I know of boards that do reasonably secure encryption and only have eight bit CPU's barely more powerful than the one in the original IBM PC keyboard. They have a special (and not too expensive) chip that implements DES, and since the original DES definition used a key that is short enough for brute force attacks nowadays, they run the message through several times with different parts of a long key. It's supposed to be safe enough to carry debit card PIN numbers under the tough European regulations. But we've got to go to nearly absurd lengths to keep that programmed-in key safe: the board is wrapped in a piece of folded paper printed with wiring patterns, then it's all potted (cast) into a block of epoxy mixed with silica grit (sand). If you take the case off, a little switch detects this and the board erases its memory in microseconds. If you somehow get past the switch and drill or cut through the epoxy, besides being darned hard on the drill bit, when you hit that paper wrapper you cut wires and the board erases. If you freeze it to weaken the epoxy and slow down the erase process, the board has a thermistor to detect falling temperatures, and erases. If you try to burn off the epoxy, that paper will go first -- and in some models, there is also a thermistor to detect rising temperatures.
As I read the claims in the patent, it sounds like they are trying to cover anything that sends one channel to a "storage medium" while watching another channel. This would seem to include my old VCR combined with a TV -- start it recording, click TV/VCR, and pick the other channel on the TV tuner. Or just set a timed record to capture Buffy and Angel while I watch something else. The good news is, that's prior art even a jury can understand. The bad news, the patent will stand as far as recording to hard drive goes unless you can either find prior art using a hard drive, or convince a court that substituting a hard drive for a tape recorder is "obvious."
We all rant and rave whenever some company tells us how we can use our product, for instance the movie industry's attempt to outlaw an open-sourced DVD player program for Linux. So it's hypocritical to then go and place restrictions on open software that amount to "Use it my way or not at all." You can never be sure that your way really is the best, but IMO even if your way really is the best, such a restriction will slow down the development of new and improved technology.
Funny how they could write that article without noticing that IQ scores in the USA have been rising steadily for about a century (ever since the first IQ test was written). It's pretty clear by now that this isn't a case of dumbing down the test, but whatever IQ tests measure really has been increasing on the average. Early explanations centered on physical factors like better food and control of diseases that may stunt the brain. But by the mid-50's there wasn't any significant room for improvement left in the physical factors, and IQ's kept on rising.
So it's got to be the more stimulating environment -- remember, at the beginning of the 20th century half of American kids grew up on farms and rarely traveled further than they could walk. Most entertainment was necessarily home made. When you got tired of listening to your sister singing the one song she knew, offkey, you could read -- most homes owned a bible and maybe one or two other books. Public libraries existed in most towns, but the 20 mile walk discouraged most rural kids. In the last 100 years, peoples', and especially childrens', horizons have widened immensely: automobiles, record players, movies, radio, airplanes, TV, etc. We are so prosperous that people can buy books for kids so young they are more likely to eat them than to look at them.
There are two observed facts that are cited against the hypothesis that the more stimulating the environment the higher the IQ. One is that many studies show a very large (over 50%) relation between IQ and genetics. Reconciling these two sets of statistics requires a very subtle relationship between IQ and genes. One theory: IQ depends mainly on stimulation. Genes determine how well you like stimulation. That is, if your genes gave you more curiousity or possibly just a little more natural ability, you will seek to learn more, and wind up considerably smarter than the average. (Remember all those stories of Abe Lincoln walking long distances to borrow a book or working hard to buy one.) TV, video games, and computers can raise the average -- but it's still the people who seek out the hardest challenges that learn the most.
The second objection to the environmental hypothesis is that for the last few decades the rise in IQ scores has not been followed by an improvement in school performance. That could mean that there is something wrong with the IQ tests, but IMO it's that the schools haven't been keeping up with the times. They still depend primarily on techniques that were developed centuries ago, and were adopted then out of necessity rather than any opinion that they were _good_. "The best school is a log with a pupil on one end and a teacher on the other", but when you have 20+ pupils, one teacher, and no technology you've just got to do the best you can. Socrates encouraged students to figure things out for themselves (circa 400BC). Teachers with larger classes had to lecture, trading a great reduction in quality for quantity. (It once also helped that the world outside school was so boring that a 70 year old man lecturing on medieval history could be stimulating by comparison.) Now computers (properly used) can free the students to go find out for themselves once again, but the education establishment is so set in it's ways that when a kid goes and learns something on his own, they resent and fear it. So the kids _are_ smarter (at least in some ways), but they're bored silly in school. Read the comments by teachers about computers in that light.
This has got nothing to do with wavelength division multiplexing. Learn a little about Fourier analysis: Only a continous, unvarying carrier wave can have a single frequency. When you modulate it to carry information, you spread out the wavelength. In the simplest modulation, on/off pulses are the sum of a whole lot of wavelengths, which travel through glass at slightly different speeds, so the pulse gets smeared out, and if not regenerated eventually the pulses overlap so much that 1's and 0's cannot be distinguished. More sophisticated modulation schemes can reduce this effect, but nothing that has been deployed outside of the laboratory eliminates it, in copper or fiber. Theoretically a pulse could be shaped so that in a non-linear medium (maybe an erbium-doped amplifier) the different wavelengths would react so as to pull the pulse back together, but so far this (called a "soliton") barely works in the laboratory. And if you know anything about telcom companies, you would know that once they get new technology working, they study it for five to twenty years before they take the risk of putting it in hard-to-reach locations.
There was one _big_ difference -- they could inductively sense the current in the cable without cutting the casing. To tap optical fibers, you've got to slit all the protective layers until you get to naked glass. Putting the casing back together so it stays watertight at high pressure is going to be difficult. Hence to tap the copper cable, they just had to send out a diver to clamp the inductive coil onto the cable, and nowadays they could just use a robot arm on the sub; for the fiber optic, you _have_ to bring the cable into a dry work-room.
I'd expect the filtering to be done by phone # or ID header. E.g., rule 2 might be "record anything addressed to Osama Bin Laden." (Rule 1 would be to watch for packets to an NSA dummy account, which would actually be new orders for the filter.)
Erbium doped fiber amplifiers only boost the light amplitude. Any data carrying signal must include a range of wavelengths, and in anything except vacuum different wavelengths travel at different rates; this is called dispersion. So the pulses spread out as they travel, and eventually you have to put in a repeater that extracts the digital data and outputs it as nicely shaped pulses again. Theoretically you could pulses called "solitons" that self-correct for dispersion, but as far as I know we're about a decade from practical applications. So there electronic repeaters out there. However, from the little I know of undersea operations, I think that unless they can steal company records pinpointing the repeater location (and I'm not sure there are any such records), you are probably better off tapping the line where you first find it than trying to follow it to find repeaters spaced a hundred miles or so apart.
As for the methods of tapping: With copper, you can just cut the outer casing, spread the wires about, and clamp an inductive pickup over each wire. You don't _have_ to penetrate the last layer of insultion, but if you want a physical splice, even this can be done without interrupting the signal. Any tap does change the impedance, which reflects a small percentage of the incoming signal, and there are (expensive) instruments that can detect this -- but if you cut in between two repeaters, you can pretty well count on that instrument not being built into the repeater. If there aren't too many wires, you might even be able to make an inductive pickup work from a few meters away.
With fiber optics, you also have to cut through individual fiber's cladding. I can't see how you could splice into a fiber optic cable without cutting the signal off entirely for seconds -- in a backbone cable, that's billions of bits gone missing, and I _hope_ a cable operator is going to notice that. But you can bend the cable until a little light starts to escape. Once again, this causes reflections and a little loss of signal strength, which an even more expensive instrument could find. But the next repeater will destroy the evidence, so if you are picking the cable off the sea bottom hundreds of miles out, the only thing that could find the tap is instruments built right into the repeaters -- and that would cost maybe $50K for each repeater, every hundred miles or so, so I don't think they'd do that.
Of course, you'd better do a _really_ good job of sealing up the cuts in the cable casing when you are done, or they'll find out about it when the cable goes bad.
On the other hand, tap the London to Paris fiber where it crosses the English channel and you probably will get caught -- probably by the Royal Navy wondering what your sub is doing, but also I'd expect the repeaters to be on dry land where the techs can run tests whenever they get nervous about the condition of the cable.
I haven't spent much time in Quebec, but your "sarcasm" may be a bit misplaced. I'm not sure how many Quebecois would choose to continue carrying the handicap of a language spoken by almost no one else in this hemisphere without these laws. I do know that when I was growing up in this part of Michigan, there were a fair number of French speakers. Now there are none, although we've still got the same proportion of people with French names. It's not quite the same situation, and I don't see it as a loss, but the people who care about preserving French in Canada aren't necessarily mistaken about coercion being the only way to do it.
Re:Article tries to make AOL look bad
on
AOL And The GPL
·
· Score: 2
just slipped through the cracks without being noticed, and will be fixed in the next release IANAL, but as I understand it
they already owe damages. The copyright notice included in the GPL license information was notification. They ignored it, and didn't meet the requirements of the license that would have allowed copying. Since real damages are hard to prove for open source, I expect this would be the "statutory damages", a fixed minimum amount, which might not be worth a big court case.
Second, the article claims the reporter called AOL and Gateway and asked about this, and received a dismissive answer. Doesn't sound like they intend to fix it until forced to.
Re:Maybe we need a standard GPL-violation Form Let
on
AOL And The GPL
·
· Score: 2
According to the article, this device is an internet terminal running Linux with: "No notices of copyright holders. No disclaimer of warranty. No source code and no directions on where and how to get the source code. No copy of the GPL, or at least none accessible or viewable from the interface provided by Gateway or America Online. No notice in the interface or help files about what is and isn't covered under the license."
If that is accurate (and I'm sure not spending $500 for a crippled laptop to find out for myself), that is 4 violations of the GPL. Maybe it's accidental -- like "I didn't know the gun was loaded" is accidental. But in a civil suit, accident is not a defense -- it just keeps you from maybe paying triple damages for malice. Furthermore, according to the article they didn't indicate any intention to correct this situation when the reporters called up to ask about it. That is, they are now getting into the situation where they provably know they are violating copyright, and keep right on doing it... IANAL, but if the copyright holders call up and get the same answer, it is definitely time to go to court.
The DMCA in the dead-tree world...
on
Digital Copyright
·
· Score: 4
By opening the cover of this book, you agree to the following license terms: [This is printed on the inside of the front cover, so you already agreed to it.] You do not own the book, you are merely licensed to read it. The publisher may revoke the license and take back the book at any time they think you might have violated the license. You may not copy the book or any portion of it allow anyone else to read the book or any portion of it, nor read it or any portion of it aloud. You may not sell the book without consent of the publisher. You may only read the book using a lamp approved by the publisher. When such lamps are withdrawn from the market, you may not read the book, even if the copyright has expired.
Of course, this won't happen with printed matter. A very long time ago some publisher put small piece of that "license" on the flyleaf of a book (just the part about re-selling), and sued some used book store; the Supreme Court threw that out, and formulated the "first sale" doctrine, that is, when you buy a book you own it. We merely have to convince them that digital publications are no different... Meanwhile, I would suggest boycotting all products that come with unreasonable restrictions, and making darn sure the store manager understands why you aren't buying.
It is a well known fact that when you become a parent you are required to retire the rational portions of your brain. Guess I must be in violation of that, I've got 2 children and 4 grandchildren and I'm not yet irrational about sex.
Why do we call entertainment which avoids any hint of the process that creates a family "family entertainment"? And why is entertainment that plays to the 15-year old mind called "adult entertainment"?
The arrest for seat-belts was a reasonable application of the Constitutional limits on federal power -- this incident should have been handled on the local level. That is, get hold of the police chief and ask what plans he has for re-educating or firing those fools working for him, and if you aren't satisfied there, work on getting a new police chief. For the Supreme Court to try to set rules on when a state or local ordinance violation requires an arrest vs a summons would be a ridiculous case of micro-management and keep them working overtime for the next 50 years, so they are lucky the Constitution does not give them the power except where illegal discrimination is clearly involved.
What does bother me is the erratic course these alleged "states rights" (in)justices are following -- it's also quite clear that under the Constitution, marijuana is not a federal matter until it crosses state lines, and how Florida counts the votes is for Florida to decide... From here, it looks like the limitations of federal powers is something they only think about when it's expedient.
Unless you are implying that one of the larger companies in the country can't hire anyone capable of running a web server. It's entirely possible that Ford can hire 10,000 lawyers easier than they can _recognize_ who's good at running a web server. Yes, it would make more sense for them to have first _asked_ for the redirect to be taken down, and second asked their webmaster if he could just block it, instead of mobilizing a battalion of lawyers. But don't expect much sense from large corporations.
One big hole with your theory is that anyone that wants to find out the real story can use whois.
So, according to your reasoning, if I say you've been convicted of child molestation, your reputation isn't injured because anyone who really cares could go down to the court house and look it up. IANAL, but I sure wouldn't go into court with that for a defense.
I do wonder what reasoning is behind filing this suit under trademark law rather than slander law, though. Also, it's plain common sense (and good manners) to ask that the re-direct be taken down before filing suit, and Ford seems to have ignored this.
except when it's in retail sales. The price of 1,000,000 chips will vary from day to day. It also varies with the size of the order and with the seller's previous experience with the buyer. Dell and IBM are dealing with variable prices everyday with the stuff they buy.
There was a time when every retail sale was negotiated -- if you wanted to buy one potato, you'd have to pick one out and argue with the grocer about the price. Some people enjoyed haggling, some hate it, but in any case you can waste a lot of time that way -- and if the grocer stands there negotiating with you for 10 minutes, he's going to have to make up the lost work time in higher prices somehow... So in the USA for a century or so retail sales have mostly been fixed price, except for items like cars and houses where there's enough money at stake to be worth it.
But computers make it easier to move back to a system where the price not only changes every few hours, but also depends on the seller's impression of the buyer, or some substitute for that. For instance, in buying an airline ticket you'll see a wide variety of prices for the same service -- the point being that if you are desperate to find _some_ flight going the right way at the right time, or are so rich that hunting through the mess isn't worth your time, the airline can get $1,000 out of you, while if you can go anytime and you put enough effort into looking up prices, you might spend $400 for the same seat.
But dynamic pricing does bother Americans, and I think it is worse when it's done by computer, since you can't haggle with it. In the old days, the ultimate argument was like "You only charged Mary Beth 3 cents for a potato this morning", and you'd either get a 3 cent potato or some (possibly valid) reason for the increase: "Her potato was smaller", or "I wasn't running out of potatoes this morning." But Amazon's computer would charge you a different price than it charged your friends, with no explanation and no one to listen to your complaints. That you could log in twice and get wildly different prices yourself just made it worse.
IBM and Dell are doing just partly dynamic pricing: changing the price as often as the price of parts or the load on their assembly line changes, but charging everyone ordering the same system at the same time the same price. It can get people a little confused, but it's not like Amazon's system. There is one thing that is crucial to keeping customers happy when you change prices frequently -- you give them a firm quotation that is good for a certain time period. If you look at Dell and the computer you want is $1,095 now, then go to other vendors to check prices, and when you come back to Dell it's now $1,195, you are quite likely to prefer the guys that have been quoting $1,200 all along. So what Dell should do is to give you a way to lock in the first price for a few days if their prices go up. I don't know if they are doing that or not -- but they'll lose customers if they don't.
And for that billboard, you could always drive over to the billboard rental agency and check their records to find out who really posted it... The point is, the average person never heard of whois, and is too lazy to use it anyhow. And anyone who thinks Ford _might_ have sponsored fuckgeneralmotors.com isn't too bright, but Ford doesn't want to lose its stupidest customers either.
Not that I'd want to leave standing any analogy between domain names and titles on billboards. In most places you _would_ get in trouble for posting "Fuck" in large letters in the public view, where it might corrupt the last virgin left in the country. 8-) In a url, only someone who already knows it can directly find it.
Ford is not trying to control who can link to them. It would be one thing (and IMO it should be legal) if there was a visible *generalmotors.com web page with a link to Ford. But instead, type *generalmotors.com and the first thing you see is a Ford corporate web site. There is nothing to say that someone else and not Ford set up that automatic transfer. Corley crossed the line here.
As I said, the big problem with relying on hard drives for backup is that a virus could rip through the whole network. Linux boxes are a little safer, two different OS's make it quite a lot safer, but not 100%. If a Windows virus is systematically eating.JPG files (for example), then a Linux server is probably safe from infection, but the.JPG files stored on it's networked drives can still be destroyed. A non-networked drive or partition on the Linux box would be safe, except that if you run a script that backs up changed files from the networked drive to the back-up drive before you know the files have been damaged...
Thanks for bringing that to my attention, it looks like it might eventually solve the problem, especially for mechanical drawings. E.g., you can export to DXF and convert DXF to SVG.
But there is a second problem: most CAD nowadays is more than just a drawing. For instance, we don't do a schematic of an electrical circuit in AutoCAD, we use a specialized tool that knows that this collection of lines is an IC, and that collection of lines is a wire. That information will be more important than the looks of the drawing when we have to re-design the board five or ten years from now. It's unlikely that SVG will provide a good way to save that information, and extremely unlikely that a different schematic drawing program will be able to import it.
I should note that the software vendors are the major part of the problem. Many of them don't _want_ your files to be portable. There is an open standard for saving schematics, and only one vendor I know of supports it (Orcad, which is one of the lowest priced ones that is capable of serious work). If one of the high-priced vendors supported it fully, customers could easily switch to Orcad. And if PADS (one of the other low-priced vendors) had been supporting a standard like this 6 or 7 years ago when their quality dropped through the floor, we could have switched to Orcad much faster. Of course, it can work the other way too: PADS seems to have re-discovered how to write a working program, but Orcad's quality is declining, so if there was an easy migration path back to PADS... But it seems like most of the vendors lack confidence that their ratio between price and quality+features is good enough for wide-open competition.
I notice that this lets one of Mundie's original statements quietly die: something like "We can make better software" [than Linux and other freeware]. To which every geek I know responded "So why don't you?" It's a plausible claim, considering that M$ has more than enough money to hire all the good OS programmers -- if they are willing to work there. But somehow they don't write reliable software. It isn't because the field is so new -- computer OS's are 40 or 50 years old!
Their most bogus claim: that freeware will suck out the money that's needed to develop high-quality commercial software. After all, most corporations are still buying Windows and MS Office because of an illusion (IMO) that it is better quality. Lousy software definitely costs more in support costs than it would take to buy good software if good software was mass-marketed.
Slashdot Mirror
How are they supposed to be compatible with previous drives??? IDE. SCSI. Or if you want to look forward instead of backwards, Firewire or USB 2.0. The article is lacking in technical details, but my best guess is that they just repackaged IDE with a custom connector, just so anyone foolish enough to buy one is locked in.
I still can't figure out why the LS-120 has failed to take off
1. Marketing. For example, the customer that told me we had to get a zip drive so they could mail us disks probably never heard of the LS-120. (Of course, now we just use CD-R and almost any computer can read the disks.
2. That compatibility comes at a price. A have had problems now and then the 1.44 meg floppies because the case is so thin it flexes more than it should and apparently the drive heads can't find the tracks. I've also had instances where the drive itself would flex too much when you bolted it in. By abandoning the standard form factors, Iomega could put massive cases on the disk and the drive, so they will never flex. The LS-120 is as thin and not-quite-rigid as the 1.44's, so unless they have compensated for this somehow, I wouldn't expect high reliability.
3. By now, the time for super-floppies has come and gone. The drives are only a little cheaper than CD-R/RW drives, the disks are a lot more expensive, and with CD-R I can create a disk that you don't need a special drive to read.
When cordless phones first became common, many people were surprised to discover that their neighbors were listening in. DUH!!!
When cellular phones came within ordinary peoples' price range, many were surprise to learn that everyone could listen in. DUH!!!
Anything you put on the radio is insecure unless it is heavily encrypted with good control of the keys. Why is that hard to understand?
The wireless keyboard and mouse could be encrypted. In fact from the article it appears that they might be encrypted; there is some sort of negotiation going on at startup, but I don't know whether that is to pick a key or simply to pick a channel. But even if the encryption is good, this live on-the-air key negotiation is a weak point. For instance, you could buy the same model of keyboard and take control after the guy turned on his computer and while he was walking over to the keyboard. Of course, you'd be entering commands blind, but there's always "del *.* (enter) y". Or since there seems to be a short list of built-in keys, you could experiment with a keyboard to find out what they all were, read the key selected from the start-up transmissions, then read out the login and password.
If you want a really secure wireless connection, then you need strong encryption with a unique key that no one else knows. Either you ship keyboard and receiver from the factory as a set (and trust the factory to erase the pre-programmed keys from their records as soon as they are used), or you have a way to temporarily bring the two devices together and connect them by a nearly untappable wire while they figure out a key.
Finally, there is a mathematical procedure that is claimed to work out a secure key by a _long_ process of exchanged messages and intensive calculation. Don't ask me to explain it. It would require enabling two-way communications, which doubles the cost of the radio circuits, and I suspect it would increase the CPU power required dramatically.
By the way, you don't need much CPU power for good secret-key encryption, you just have to design right. I know of boards that do reasonably secure encryption and only have eight bit CPU's barely more powerful than the one in the original IBM PC keyboard. They have a special (and not too expensive) chip that implements DES, and since the original DES definition used a key that is short enough for brute force attacks nowadays, they run the message through several times with different parts of a long key. It's supposed to be safe enough to carry debit card PIN numbers under the tough European regulations. But we've got to go to nearly absurd lengths to keep that programmed-in key safe: the board is wrapped in a piece of folded paper printed with wiring patterns, then it's all potted (cast) into a block of epoxy mixed with silica grit (sand). If you take the case off, a little switch detects this and the board erases its memory in microseconds. If you somehow get past the switch and drill or cut through the epoxy, besides being darned hard on the drill bit, when you hit that paper wrapper you cut wires and the board erases. If you freeze it to weaken the epoxy and slow down the erase process, the board has a thermistor to detect falling temperatures, and erases. If you try to burn off the epoxy, that paper will go first -- and in some models, there is also a thermistor to detect rising temperatures.
As I read the claims in the patent, it sounds like they are trying to cover anything that sends one channel to a "storage medium" while watching another channel. This would seem to include my old VCR combined with a TV -- start it recording, click TV/VCR, and pick the other channel on the TV tuner. Or just set a timed record to capture Buffy and Angel while I watch something else. The good news is, that's prior art even a jury can understand. The bad news, the patent will stand as far as recording to hard drive goes unless you can either find prior art using a hard drive, or convince a court that substituting a hard drive for a tape recorder is "obvious."
We all rant and rave whenever some company tells us how we can use our product, for instance the movie industry's attempt to outlaw an open-sourced DVD player program for Linux. So it's hypocritical to then go and place restrictions on open software that amount to "Use it my way or not at all." You can never be sure that your way really is the best, but IMO even if your way really is the best, such a restriction will slow down the development of new and improved technology.
Funny how they could write that article without noticing that IQ scores in the USA have been rising steadily for about a century (ever since the first IQ test was written). It's pretty clear by now that this isn't a case of dumbing down the test, but whatever IQ tests measure really has been increasing on the average. Early explanations centered on physical factors like better food and control of diseases that may stunt the brain. But by the mid-50's there wasn't any significant room for improvement left in the physical factors, and IQ's kept on rising.
So it's got to be the more stimulating environment -- remember, at the beginning of the 20th century half of American kids grew up on farms and rarely traveled further than they could walk. Most entertainment was necessarily home made. When you got tired of listening to your sister singing the one song she knew, offkey, you could read -- most homes owned a bible and maybe one or two other books. Public libraries existed in most towns, but the 20 mile walk discouraged most rural kids. In the last 100 years, peoples', and especially childrens', horizons have widened immensely: automobiles, record players, movies, radio, airplanes, TV, etc. We are so prosperous that people can buy books for kids so young they are more likely to eat them than to look at them.
There are two observed facts that are cited against the hypothesis that the more stimulating the environment the higher the IQ. One is that many studies show a very large (over 50%) relation between IQ and genetics. Reconciling these two sets of statistics requires a very subtle relationship between IQ and genes. One theory: IQ depends mainly on stimulation. Genes determine how well you like stimulation. That is, if your genes gave you more curiousity or possibly just a little more natural ability, you will seek to learn more, and wind up considerably smarter than the average. (Remember all those stories of Abe Lincoln walking long distances to borrow a book or working hard to buy one.) TV, video games, and computers can raise the average -- but it's still the people who seek out the hardest challenges that learn the most.
The second objection to the environmental hypothesis is that for the last few decades the rise in IQ scores has not been followed by an improvement in school performance. That could mean that there is something wrong with the IQ tests, but IMO it's that the schools haven't been keeping up with the times. They still depend primarily on techniques that were developed centuries ago, and were adopted then out of necessity rather than any opinion that they were _good_. "The best school is a log with a pupil on one end and a teacher on the other", but when you have 20+ pupils, one teacher, and no technology you've just got to do the best you can. Socrates encouraged students to figure things out for themselves (circa 400BC). Teachers with larger classes had to lecture, trading a great reduction in quality for quantity. (It once also helped that the world outside school was so boring that a 70 year old man lecturing on medieval history could be stimulating by comparison.) Now computers (properly used) can free the students to go find out for themselves once again, but the education establishment is so set in it's ways that when a kid goes and learns something on his own, they resent and fear it. So the kids _are_ smarter (at least in some ways), but they're bored silly in school. Read the comments by teachers about computers in that light.
This has got nothing to do with wavelength division multiplexing. Learn a little about Fourier analysis: Only a continous, unvarying carrier wave can have a single frequency. When you modulate it to carry information, you spread out the wavelength. In the simplest modulation, on/off pulses are the sum of a whole lot of wavelengths, which travel through glass at slightly different speeds, so the pulse gets smeared out, and if not regenerated eventually the pulses overlap so much that 1's and 0's cannot be distinguished. More sophisticated modulation schemes can reduce this effect, but nothing that has been deployed outside of the laboratory eliminates it, in copper or fiber. Theoretically a pulse could be shaped so that in a non-linear medium (maybe an erbium-doped amplifier) the different wavelengths would react so as to pull the pulse back together, but so far this (called a "soliton") barely works in the laboratory. And if you know anything about telcom companies, you would know that once they get new technology working, they study it for five to twenty years before they take the risk of putting it in hard-to-reach locations.
There was one _big_ difference -- they could inductively sense the current in the cable without cutting the casing. To tap optical fibers, you've got to slit all the protective layers until you get to naked glass. Putting the casing back together so it stays watertight at high pressure is going to be difficult. Hence to tap the copper cable, they just had to send out a diver to clamp the inductive coil onto the cable, and nowadays they could just use a robot arm on the sub; for the fiber optic, you _have_ to bring the cable into a dry work-room.
I'd expect the filtering to be done by phone # or ID header. E.g., rule 2 might be "record anything addressed to Osama Bin Laden." (Rule 1 would be to watch for packets to an NSA dummy account, which would actually be new orders for the filter.)
Erbium doped fiber amplifiers only boost the light amplitude. Any data carrying signal must include a range of wavelengths, and in anything except vacuum different wavelengths travel at different rates; this is called dispersion. So the pulses spread out as they travel, and eventually you have to put in a repeater that extracts the digital data and outputs it as nicely shaped pulses again. Theoretically you could pulses called "solitons" that self-correct for dispersion, but as far as I know we're about a decade from practical applications. So there electronic repeaters out there. However, from the little I know of undersea operations, I think that unless they can steal company records pinpointing the repeater location (and I'm not sure there are any such records), you are probably better off tapping the line where you first find it than trying to follow it to find repeaters spaced a hundred miles or so apart.
As for the methods of tapping: With copper, you can just cut the outer casing, spread the wires about, and clamp an inductive pickup over each wire. You don't _have_ to penetrate the last layer of insultion, but if you want a physical splice, even this can be done without interrupting the signal. Any tap does change the impedance, which reflects a small percentage of the incoming signal, and there are (expensive) instruments that can detect this -- but if you cut in between two repeaters, you can pretty well count on that instrument not being built into the repeater. If there aren't too many wires, you might even be able to make an inductive pickup work from a few meters away.
With fiber optics, you also have to cut through individual fiber's cladding. I can't see how you could splice into a fiber optic cable without cutting the signal off entirely for seconds -- in a backbone cable, that's billions of bits gone missing, and I _hope_ a cable operator is going to notice that. But you can bend the cable until a little light starts to escape. Once again, this causes reflections and a little loss of signal strength, which an even more expensive instrument could find. But the next repeater will destroy the evidence, so if you are picking the cable off the sea bottom hundreds of miles out, the only thing that could find the tap is instruments built right into the repeaters -- and that would cost maybe $50K for each repeater, every hundred miles or so, so I don't think they'd do that. Of course, you'd better do a _really_ good job of sealing up the cuts in the cable casing when you are done, or they'll find out about it when the cable goes bad.
On the other hand, tap the London to Paris fiber where it crosses the English channel and you probably will get caught -- probably by the Royal Navy wondering what your sub is doing, but also I'd expect the repeaters to be on dry land where the techs can run tests whenever they get nervous about the condition of the cable.
I haven't spent much time in Quebec, but your "sarcasm" may be a bit misplaced. I'm not sure how many Quebecois would choose to continue carrying the handicap of a language spoken by almost no one else in this hemisphere without these laws. I do know that when I was growing up in this part of Michigan, there were a fair number of French speakers. Now there are none, although we've still got the same proportion of people with French names. It's not quite the same situation, and I don't see it as a loss, but the people who care about preserving French in Canada aren't necessarily mistaken about coercion being the only way to do it.
just slipped through the cracks without being noticed, and will be fixed in the next release IANAL, but as I understand it they already owe damages. The copyright notice included in the GPL license information was notification. They ignored it, and didn't meet the requirements of the license that would have allowed copying. Since real damages are hard to prove for open source, I expect this would be the "statutory damages", a fixed minimum amount, which might not be worth a big court case.
Second, the article claims the reporter called AOL and Gateway and asked about this, and received a dismissive answer. Doesn't sound like they intend to fix it until forced to.
According to the article, this device is an internet terminal running Linux with: "No notices of copyright holders. No disclaimer of warranty. No source code and no directions on where and how to get the source code. No copy of the GPL, or at least none accessible or viewable from the interface provided by Gateway or America Online. No notice in the interface or help files about what is and isn't covered under the license."
If that is accurate (and I'm sure not spending $500 for a crippled laptop to find out for myself), that is 4 violations of the GPL. Maybe it's accidental -- like "I didn't know the gun was loaded" is accidental. But in a civil suit, accident is not a defense -- it just keeps you from maybe paying triple damages for malice. Furthermore, according to the article they didn't indicate any intention to correct this situation when the reporters called up to ask about it. That is, they are now getting into the situation where they provably know they are violating copyright, and keep right on doing it... IANAL, but if the copyright holders call up and get the same answer, it is definitely time to go to court.
By opening the cover of this book, you agree to the following license terms: [This is printed on the inside of the front cover, so you already agreed to it.] You do not own the book, you are merely licensed to read it. The publisher may revoke the license and take back the book at any time they think you might have violated the license. You may not copy the book or any portion of it allow anyone else to read the book or any portion of it, nor read it or any portion of it aloud. You may not sell the book without consent of the publisher. You may only read the book using a lamp approved by the publisher. When such lamps are withdrawn from the market, you may not read the book, even if the copyright has expired.
Of course, this won't happen with printed matter. A very long time ago some publisher put small piece of that "license" on the flyleaf of a book (just the part about re-selling), and sued some used book store; the Supreme Court threw that out, and formulated the "first sale" doctrine, that is, when you buy a book you own it. We merely have to convince them that digital publications are no different... Meanwhile, I would suggest boycotting all products that come with unreasonable restrictions, and making darn sure the store manager understands why you aren't buying.
And why would any adult be willing to run for Congress?
It is a well known fact that when you become a parent you are required to retire the rational portions of your brain. Guess I must be in violation of that, I've got 2 children and 4 grandchildren and I'm not yet irrational about sex.
Why do we call entertainment which avoids any hint of the process that creates a family "family entertainment"? And why is entertainment that plays to the 15-year old mind called "adult entertainment"?
The arrest for seat-belts was a reasonable application of the Constitutional limits on federal power -- this incident should have been handled on the local level. That is, get hold of the police chief and ask what plans he has for re-educating or firing those fools working for him, and if you aren't satisfied there, work on getting a new police chief. For the Supreme Court to try to set rules on when a state or local ordinance violation requires an arrest vs a summons would be a ridiculous case of micro-management and keep them working overtime for the next 50 years, so they are lucky the Constitution does not give them the power except where illegal discrimination is clearly involved.
What does bother me is the erratic course these alleged "states rights" (in)justices are following -- it's also quite clear that under the Constitution, marijuana is not a federal matter until it crosses state lines, and how Florida counts the votes is for Florida to decide... From here, it looks like the limitations of federal powers is something they only think about when it's expedient.
Unless you are implying that one of the larger companies in the country can't hire anyone capable of running a web server. It's entirely possible that Ford can hire 10,000 lawyers easier than they can _recognize_ who's good at running a web server. Yes, it would make more sense for them to have first _asked_ for the redirect to be taken down, and second asked their webmaster if he could just block it, instead of mobilizing a battalion of lawyers. But don't expect much sense from large corporations.
One big hole with your theory is that anyone that wants to find out the real story can use whois. So, according to your reasoning, if I say you've been convicted of child molestation, your reputation isn't injured because anyone who really cares could go down to the court house and look it up. IANAL, but I sure wouldn't go into court with that for a defense.
I do wonder what reasoning is behind filing this suit under trademark law rather than slander law, though. Also, it's plain common sense (and good manners) to ask that the re-direct be taken down before filing suit, and Ford seems to have ignored this.
except when it's in retail sales. The price of 1,000,000 chips will vary from day to day. It also varies with the size of the order and with the seller's previous experience with the buyer. Dell and IBM are dealing with variable prices everyday with the stuff they buy. There was a time when every retail sale was negotiated -- if you wanted to buy one potato, you'd have to pick one out and argue with the grocer about the price. Some people enjoyed haggling, some hate it, but in any case you can waste a lot of time that way -- and if the grocer stands there negotiating with you for 10 minutes, he's going to have to make up the lost work time in higher prices somehow... So in the USA for a century or so retail sales have mostly been fixed price, except for items like cars and houses where there's enough money at stake to be worth it.
But computers make it easier to move back to a system where the price not only changes every few hours, but also depends on the seller's impression of the buyer, or some substitute for that. For instance, in buying an airline ticket you'll see a wide variety of prices for the same service -- the point being that if you are desperate to find _some_ flight going the right way at the right time, or are so rich that hunting through the mess isn't worth your time, the airline can get $1,000 out of you, while if you can go anytime and you put enough effort into looking up prices, you might spend $400 for the same seat.
But dynamic pricing does bother Americans, and I think it is worse when it's done by computer, since you can't haggle with it. In the old days, the ultimate argument was like "You only charged Mary Beth 3 cents for a potato this morning", and you'd either get a 3 cent potato or some (possibly valid) reason for the increase: "Her potato was smaller", or "I wasn't running out of potatoes this morning." But Amazon's computer would charge you a different price than it charged your friends, with no explanation and no one to listen to your complaints. That you could log in twice and get wildly different prices yourself just made it worse.
IBM and Dell are doing just partly dynamic pricing: changing the price as often as the price of parts or the load on their assembly line changes, but charging everyone ordering the same system at the same time the same price. It can get people a little confused, but it's not like Amazon's system. There is one thing that is crucial to keeping customers happy when you change prices frequently -- you give them a firm quotation that is good for a certain time period. If you look at Dell and the computer you want is $1,095 now, then go to other vendors to check prices, and when you come back to Dell it's now $1,195, you are quite likely to prefer the guys that have been quoting $1,200 all along. So what Dell should do is to give you a way to lock in the first price for a few days if their prices go up. I don't know if they are doing that or not -- but they'll lose customers if they don't.
And for that billboard, you could always drive over to the billboard rental agency and check their records to find out who really posted it... The point is, the average person never heard of whois, and is too lazy to use it anyhow. And anyone who thinks Ford _might_ have sponsored fuckgeneralmotors.com isn't too bright, but Ford doesn't want to lose its stupidest customers either.
Not that I'd want to leave standing any analogy between domain names and titles on billboards. In most places you _would_ get in trouble for posting "Fuck" in large letters in the public view, where it might corrupt the last virgin left in the country. 8-) In a url, only someone who already knows it can directly find it.
Ford is not trying to control who can link to them. It would be one thing (and IMO it should be legal) if there was a visible *generalmotors.com web page with a link to Ford. But instead, type *generalmotors.com and the first thing you see is a Ford corporate web site. There is nothing to say that someone else and not Ford set up that automatic transfer. Corley crossed the line here.
As I said, the big problem with relying on hard drives for backup is that a virus could rip through the whole network. Linux boxes are a little safer, two different OS's make it quite a lot safer, but not 100%. If a Windows virus is systematically eating .JPG files (for example), then a Linux server is probably safe from infection, but the .JPG files stored on it's networked drives can still be destroyed. A non-networked drive or partition on the Linux box would be safe, except that if you run a script that backs up changed files from the networked drive to the back-up drive before you know the files have been damaged...
Thanks for bringing that to my attention, it looks like it might eventually solve the problem, especially for mechanical drawings. E.g., you can export to DXF and convert DXF to SVG.
But there is a second problem: most CAD nowadays is more than just a drawing. For instance, we don't do a schematic of an electrical circuit in AutoCAD, we use a specialized tool that knows that this collection of lines is an IC, and that collection of lines is a wire. That information will be more important than the looks of the drawing when we have to re-design the board five or ten years from now. It's unlikely that SVG will provide a good way to save that information, and extremely unlikely that a different schematic drawing program will be able to import it.
I should note that the software vendors are the major part of the problem. Many of them don't _want_ your files to be portable. There is an open standard for saving schematics, and only one vendor I know of supports it (Orcad, which is one of the lowest priced ones that is capable of serious work). If one of the high-priced vendors supported it fully, customers could easily switch to Orcad. And if PADS (one of the other low-priced vendors) had been supporting a standard like this 6 or 7 years ago when their quality dropped through the floor, we could have switched to Orcad much faster. Of course, it can work the other way too: PADS seems to have re-discovered how to write a working program, but Orcad's quality is declining, so if there was an easy migration path back to PADS... But it seems like most of the vendors lack confidence that their ratio between price and quality+features is good enough for wide-open competition.
I notice that this lets one of Mundie's original statements quietly die: something like "We can make better software" [than Linux and other freeware]. To which every geek I know responded "So why don't you?" It's a plausible claim, considering that M$ has more than enough money to hire all the good OS programmers -- if they are willing to work there. But somehow they don't write reliable software. It isn't because the field is so new -- computer OS's are 40 or 50 years old!
Their most bogus claim: that freeware will suck out the money that's needed to develop high-quality commercial software. After all, most corporations are still buying Windows and MS Office because of an illusion (IMO) that it is better quality. Lousy software definitely costs more in support costs than it would take to buy good software if good software was mass-marketed.