[Decimal powers are] patently incorrect, and originated as false advertising among hard drive producers....
kilo- was defined as 1,000 well before some computer scientist decided that 2.4% was close enough for a convenience. For most measurements (including mail quota), the binary definitions aren't convenient and thus aren't used; when they are used, the preferred (ISO, IEC, IEEE, NIST, CIPM-endorsed) prefixes are kibi-, mebi-, gibi-, etc.
The GPL allows you to change the GPL_ONLY symbols anyway
Linus addressed this during the original discussion. The EXPORT_SYMBOL_GPL() macro means nothing by itself; it merely marks the symbols that can be used only by GPL-compatible modules. Removing this marking doesn't change the nature of the symbols.
(even tho Alan Cox was threatening to put people in jail under the DMCA if they did so)
If your proprietary module is legal, you can always argue "interoperability", falsify your licence string and call the GPL_ONLY stuff legally. This sort of thing has been held up in court many times.
If I could convince a court that my module wasn't a derivative work, sure. You seem to have misunderstood the entire point of the MODULE_LICENSE dance.
In short, it would have been a lot easier if they had just put in a textfile saying "Please don't call these functions!" rather than this complex symbol/licence crap which doesn't really do anything.
This way ensures that any developer using GPL_ONLY symbols is aware of their status.
I thought it was patents that you could selectivly defend... and copyrights and trademarks were the ones that needed to be vigorously defended to remain current...?
No, it's pretty much just trademarks. Copyright and patent suits may be rejected due to laches (unconscionable delay in bringing suit) or estoppel (leading the other party to believe one wouldn't enforce one's rights), but that really applies to only a particular infringer--i.e., just because someone else is doing it doesn't mean you can too.
None of this is relevant to Linux kernel binary modules, though, because the message always has been the same: derivative works must be GPL; nonderivative works are outside the scope of the kernel developers' copyrights.
Incidentally, I don't recall Eolas ever saying that it would stop with Microsoft--Doyle posed that as a hypothetical scenario. It's irrelevant, of course, since the patent was invalidated.
But how can some symbols be GPL and some not, considering that, as it stands, the entirety of the core kernel code is licensed under the GPL, and the GPL does not allow exceptions to that licensing?
Not all of the interfaces to the Linux kernel are unique to Linux; some are decades old and more or less standard across anything vaguely Unix. If a bit of code wasn't written for Linux, it can't be considered a derivative work, and the GPL doesn't apply.
Linus has said that he considers binary modules to not be far enough removed from GPL code and thus infringing, but since binary modules have been around since very early on in the kernels development history without any enforcement of the GPL with regards to them, wouldnt that potentially count against the GPL applying to binary modules if someone did decide to take action? Doesnt the whole idea of kernel license strings interfere with this view as well? If modules are infringing if they arent GPL, then why would they need to tell the kernel that they arent under the GPL?
What Linus has said is that modules can't avoid the GPL simply by being modules. At one time, the interfaces were limited enough that no modules were considered derivative, but that has changed. A kernel developer now can choose to EXPORT_SYMBOL() or EXPORT_SYMBOL_GPL() to indicate whether using a given symbol constitutes derivation; the MODULE_LICENSE string is simply a way to enforce developer awareness.
Also, where in the Kernel license does it require you to be truthful to the kernel about your modules license? Nowhere, because it cant. The GPL will not allow you to put that limitation on use of the kernel.
It's a moot point; a proprietary module that uses GPL symbols is an unauthorized derivative.
Perhaps. I've explored that line of thought before, and murder fits the definition better than does copyright infringement. That isn't how the law works, though; and as above, if he wants to change that, he should start by providing a new definition.
One of the big problem that Dell encountered is that 50%+ of the Linux customers wanted Distro X instead of RedHat. It would be a configuration nightmare to install and support RedHat (2 versions), SuSE, Mandrake, etc etc etc.
I'm sure there was some of that; I'm skeptical about the percentage, but it doesn't diminish my points. In fact, it's illustrative--someone wanting a different distro was better served by configuring a system with Windows.
The other issue is that adding products into a sales channel is not as trivial as it might seem. It looked like they had to tack the Linux section onto the side of their website because their existing templates weren't ready for that sort product.
The OS was just another drop-down box on the build-to-order page. The Linux pages had to be tacked on only because of the limited configurations, and Dell managed to link more visibly to other tacked-on pages.
Which is why Dell now offers "bare" n-series desktops. You BYOOS and arrange the support yourself.
Yep, and they can be built to order, too. Still only select models, though.
Many years ago (early 1900's, I think), book publishers tried to insert conditions on the front page of a book saying that you couldn't resell the book, etc. It was struck down, and thus the "doctrine of first sale" was born.
The doctrine of first sale was implicit in earlier decisions, but the most famous precedent was set in 1908 in Bobbs-Merrill v. Strauss. The Bobbs-Merrill Company published a novel containing below the copyright notice an assertion that any sale for less than $1 would constitute copyright infringement. As you know, the court didn't buy it.
OK... try rewiring the cable coming into your house to get HBO/Showtime/Whatever. Don't think "theft of service" is going to apply?
US Code calls it 'unauthorized reception'.
Let's say it cost you several thousand to generate some document that you've got on your hard drive and I hack into the box and get it. Is it now OK because I was smart enough to get it and you weren't smart enough to protect it?
Nope. Hacking into the box is illegal. Copying the document is illegal. There are a lot of things that are illegal and aren't theft. Jaywalking. Murder.
At least be mature enough to admit that what you're doing is no different (legally or morally) from stealing tangible goods.
THEY ARE THE SAME. It denies the victim revenue.
We can debate morals, if you'd like, but the law is clear: copyright infringement and theft are distinct offenses. If you think they should be conflated, start by proposing a new definition for the combined law. You'll have to do better than 'denying revenue', though--fair use, repair, buying from a competitor, and doing without all have that effect.
Dell temporarily sold their home desktop line with Red Hat 6.x preloaded. IBM Also sold Thinkpads with Red Hat 6.x preloaded. Both got canceled due to poor sales.
Dell also sold laptops with Red Hat; like the desktops, only a select few configurations were available, and each cost the same as the equivalent Windows configuration. The Linuxcare contracts might have justified the price on Dell's end, but they weren't of much use to customers, especially when getting various parts working correctly required unsupported software. Finally, as an anonymous sibling noted, Dell hardly advertised these offerings, even on its own website.
Obviously, producing compelling products and informing its customers would have cost Dell more, maybe more than it would have brought in. However, mediocre products poorly marketed tend not to sell regardless of the target market.
That's more true than you let on; the GPL was written in reaction to restrictions imposed on works through copyright. It depends on the same copyright and contract law as proprietary software licenses by design, ensuring that any attempts to weaken the GPL would have the same effect on all software licenses. Some GPL advocates miss the forest for the trees, but RMS (for one) has no particular attachment to the GPL per se; it is a means to an end, a clever hack which has the desired effect within the current system. The principles of free software are found in the four freedoms, not the language of the GPL.
Which is why the winnowing/chaffing approach is so interesting, as one could arrange it so a "fake" key could decrypt to innocuous information (e.g. "last 5 seconds was 30 mph, with brake lightly applied").
I know of the technique, but it frankly it seems completely detached from the topic of the article. There's little reason to encrypt the data--sign, yes, encrypt, no. An encrypting data recorder would most likely use symmetric crypto with the keys disclosed to law enforcement agencies, insurance companies, and probably not the owner. I doubt there's a significant market for chaffing public-key automobile data recorders.
What do you think would (or should) happen if the black box information was encrypted in such a way that only the car owner could decrypt it (e.g. with the owner's public key, requiring the owner's private key to decrypt)? A court order to demand the decryption key?
The court would issue a subpoena for the recording, and the owner could be held in contempt for refusing to produce the key. (What happens if the owner claims to have forgotten his passphrase is, as far as I know, an open question.)
The box may be purely evidence, though I am unconvinced.
What else do you think it is?
In any case, search warrants overcome 4th amendment protections. There is no warrant that can compel a person to testify against himself.
We're not talking about a person testifying against himself; we're talking about property, to which the Fifth Amendment does not extend. I offered search warrants as an illustration of that fact. Searching a vehicle used as a homicide weapon is well within the bounds of the Fourth Amendment.
There are several different kinds of fluorocarbons. Chlorofluorocarbons (CFCs) deplete stratospheric ozone; perfluorocarbons (PFCs) like Fluorinert have high greenhouse potentials but are not dangerous to the ozone layer.
The counterargument to my counterargument is that by burning & re-ripping you are losing quality, but the counterargument to this counterargument of my counterargument is that if you were enough of an audiophile to care about this, you wouldn't be buying 128K mp4s from iTMS anyway.
Burning and reripping is time-consuming and wasteful; the right software can eliminate the waste and some of the time, but even that is flirting with the DMCA. Also, iTMS files represent a minimum acceptable level of quality to some people; settling for 128 Kbps AAC doesn't imply a total lack of standards.
Slashdot Mirror
The Unistroke patent was filed as a continuation of an application from 1993 and thus inherited the earlier filing date.
None of this is relevant to Linux kernel binary modules, though, because the message always has been the same: derivative works must be GPL; nonderivative works are outside the scope of the kernel developers' copyrights.
Incidentally, I don't recall Eolas ever saying that it would stop with Microsoft--Doyle posed that as a hypothetical scenario. It's irrelevant, of course, since the patent was invalidated.
It's a moot point; a proprietary module that uses GPL symbols is an unauthorized derivative.
Nope. Hacking into the box is illegal. Copying the document is illegal. There are a lot of things that are illegal and aren't theft. Jaywalking. Murder.
We can debate morals, if you'd like, but the law is clear: copyright infringement and theft are distinct offenses. If you think they should be conflated, start by proposing a new definition for the combined law. You'll have to do better than 'denying revenue', though--fair use, repair, buying from a competitor, and doing without all have that effect.
Obviously, producing compelling products and informing its customers would have cost Dell more, maybe more than it would have brought in. However, mediocre products poorly marketed tend not to sell regardless of the target market.
There are several different kinds of fluorocarbons. Chlorofluorocarbons (CFCs) deplete stratospheric ozone; perfluorocarbons (PFCs) like Fluorinert have high greenhouse potentials but are not dangerous to the ozone layer.
Burning and reripping is time-consuming and wasteful; the right software can eliminate the waste and some of the time, but even that is flirting with the DMCA. Also, iTMS files represent a minimum acceptable level of quality to some people; settling for 128 Kbps AAC doesn't imply a total lack of standards.