I thought they put that in as an Easter egg... the Patriot Act isn't a novel. Though some Eastern bloc countries allegedly used 1984 as a HOWTO, or a specification of an ideal government.
That's pretty much the definition of a vanity press. PublishAmerica (who were nearly duped into publishing Atlanta Nights) are just a bit better known than most.
No one will admit to running a vanity press, which is why they have to be exposed, and naive authors warned against them.
There are legitimate companies where you can pay to have your book published, but these guys are up-front about what they offer and how much it costs. They don't pretend to be choosy about what they print. (400 pages of/dev/urandom? No problem.) They also tend to be realistic about your chances of having a bestseller, or even turning a profit (comparable to the odds of finding Hamlet's soliloquy in those 400 pages).
I agree that the committee's ideas about what to fund seem strange sometimes, though the "strange" Good Causes will get more media attention than the "normal" ones. Most of the money paid into the lottery goes straight back out to those who buy tickets, in the form of prizes.
The money spent on tickets is distributed as follows:
28% isn't exactly a very small percentage. Though who knows - maybe the Government's 12% should go to the Good Causes fund too. Maybe Branson could run it more efficiently than Camelot, but if the Government won't give up its share, a Branson-run lottery would give between 28% and 33% of the sales to good causes. That doesn't sound like so much of an improvement to me. And that's assuming ticket sales don't go down when he takes over.
I know the plural of "anecdote" is not "data", but Firefox 2 has been very stable for me. This is on Mandriva 2007. The only extension I use is NoScript.
[steve@shamrock ~]$ ps -ef|grep fire steve 6438 6355 0 Feb22 pts/0 00:00:00/bin/sh/usr/bin/firefox steve 6446 6438 0 Feb22 pts/0 00:00:00/bin/sh/usr/lib/mozilla-firefox-2.0.0.11/run-mozilla.sh/usr/lib/mozilla-firefox-2.0.0.11/mozilla-firefox-bin steve 6451 6446 0 Feb22 pts/0 05:30:34/usr/lib/mozilla-firefox-2.0.0.11/mozilla-firefox-bin steve 29427 6355 0 15:55 pts/0 00:00:00 grep --color fire
Mozilla 1.x would consistently crash after a week or two, with an X error that made me think it had run out of some kind of resource.
I agree Firefox 2 is a bit of a memory hog. Mine is using about 350Mb at the moment, with 8 tabs open, though most of that is swapped out. Maybe I'm getting old, but it feels fast enough. It sounds as though version 3 will be a pleasant surprise.
I live in the UK, and our minister in charge of education really is called Mr Balls. I have to remind myself not to snigger every time I hear about him.
"Happy slapping" is the act of assaulting someone while one of your friends films it, usually on a camera phone, the aim being to send the recording to all your other friends to prove how tough you are (or something like that)...
That would probably be The Metamorphosis of Prime Intellect, by the same localroger chappie. (Too lazy to dig out a link to it.)
I read it a couple of years ago and, while the plot was interesting, I wouldn't say I enjoyed it very much. I'm not keen on blood and gore, and I think the large amount of it in the first half of the book put me off.
There's an urban legend of a Superman costume with the lable "Warning, does not enable user to fly."
Probably it is an urban legend, but I thought that one was fairly sensible, given who's likely to be wearing the costume. Consider: you're four years old and like watching Superman on TV. You've figured out that Superman and Clark Kent are really the same person, but for some reason, Clark never outruns (outflies?) a speeding bullet or leaps tall buildings in a single bound. What's the difference? Clark just wears a boring blue suit like millions of other guys, but Superman wears this cool costume with the underpants outside. Obviously, therefore, the costume must be the source of Superman's powers... [*]
[*] I didn't believe this when I was four. I read the comics, where you can see that Clark has the Superman costume underneath his suit. Besides, you couldn't buy superhero costumes in those days, and I must have known that one I improvised wouldn't have the special powers of the real thing...
D'OH! Ever think you read something, but you didn't?
More often than I care to remember:-)
Well, the patent office isn't doing it -- and if it's hurting the public, maybe the public has to do something about it?
I can't really argue with that. If someone is attacking you, you'd be foolish not to defend yourself. (Or run away!) But surely it's better if you can prevent them from being allowed to attack you in the first place. Both approaches can proceed in parallel, of course. It just makes me angry that people such as those at WikiPatents have to spend so much time and effort cleaning up other people's incompetence and malice, when they could be making positive contributions to the useful arts and sciences.
[M]ost CIOs aren't taking the "It's not proven legally" position -- they're influenced by the FUD.
I fear you're right there. I try to think with my head more than my gut. That's probably one reason I'm a software engineer, not a CIO:-) "Prove it or shut up" works when your adversary is the prosecution in a criminal trial, or a scientist advocating a new theory. But I suppose when your opponent is shovelling great steaming piles of FUD onto you or your (potential) customers, you have to do something to clear it away. I'd love to be a fly on the wall when the first CIO asks a Microsoft sales rep, "So, which of your patents does Linux infringe on, then?"
Good points... if you want to continue the metaphor, remember that Britain had an empire in 1914 too... the biggest the world had ever seen. (The saying "The sun never sets on the British Empire" meant literally that - whatever time of day or night it was in London, it would be daytime for some part of the Empire.) Nowadays, our possessions consist of Northern Ireland and a few islands that function mainly as tax havens and tourist traps.
I can't really see a big company starting a patent war, unless they think they have no alternative. Suppose Microsoft were to sue Red Hat, claiming that Red Hat Linux infringes one or more Microsoft patents. Before the ink was dry on the writs, IBM would be telling Microsoft to drop the suit or face a suit for infringement of (say) several hundred IBM patents. The thing is, Microsoft knows that would (almost certainly) happen, so they're not going to sue Red Hat unless they can be very sure of winning against IBM.
Here, as before World War I, you have an enormous buildup of armed forces on all sides. But I think the reason World War I escalated so quickly and became so devastating was alliances between countries and, worse, secret alliances. Two or more countries would promise to help one another if either was attacked... but they wouldn't tell anyone else about it. If several countries enter such treaties, that pretty much guarantees that any war will escalate until it involves everybody.
The laws governing companies make it difficult (in theory) for publicly-traded companies to form secret alliances. And patents are all visible to everybody, so someone who's thinking of starting a patent war can see what the opposition could decide to use against them.
I don't want a patent war, but it might be the only way to make lawmakers see how broken the patent system has become. It seems humanity never learns a lesson except the hard way...
I'm aware of the triple damages rule. It was the last bullet point in my list.
There's a wiki about patents, called, appropriately enough, WikiPatents. I don't know much more about them; I just found them linked from pubpat.org today. They're looking at all US patents, not just those related to software. What's on the site might not be admissible in a court case, but it would certainly give a head start to a lawyer who's trying to get a patent invalidated.
Worthy though such projects are, they're treating the symptoms, not the disease. Why the hell should the public have to do the patent office's job for it? The real solution is to stop bad patents from being issued in the first place.
But really, as this comment points out, since Microsoft have made the allegation, it's up to them to prove it. The free software community doesn't have to disprove it. We don't have to do anything beyond saying (repeatedly), "Tell us which features of which projects infringe which of your patents. Then we will get those patents invalidated, or work around them, or remove the infringing code, in that order. Until you give us the information we have asked for, we will have to assume that either there is no infringement, or you do not want the infringement cured. Both alternatives imply you are acting in bad faith."
The article author is conflating patents and copyright. Is it too much to ask that someone who (presumably) gets paid to write this stuff would know the difference? From the summary:
After all, Microsoft does hold a lot of patents and while Linux is open source and we can all take a look at the source code, only Microsoft has access to most of its source code so it isn't all that difficult for it to prove - to itself at any rate - that there are IP infringements contained in Linux.
Having access to Microsoft's source code would neither help nor hinder anyone in deciding whether Linux infringes any Microsoft patents. The whole point of patents is that they're visible to the public. The US government even maintains a database of them that anyone with a web browser can look at! A patent holder doesn't have to implement the invention that their patent describes before they can sue someone else for infringing it. If they did, patent trolls couldn't exist.
Anyone could drive a stake through the heart of Microsoft's FUD campaign by getting a list of all patents held by Microsoft and proving, for each patent, either (a) Linux does not infringe the patent or (b) the patent is invalid. I suspect this will not happen, for any or all of the following reasons:
Nobody who understands both patent law and the internals of Linux will come forward.
Nobody will want to spend the time or money to do it.
The open-source community as a whole believes the risk of Microsoft actually suing anybody is negligible. As others have pointed out, FUD is much cheaper than lawsuits, and probably much more effective at detering businesses from moving to Linux.
The deliberately vague and obfuscated language in which patents are written means that the question of whether something infringes a patent is not black-and-white. Sometimes, the only way to be sure is to put the matter in front of a judge and jury. (I'd really like to see a judge rule a patent to be invalid because a person with ordinary skill in the art wouldn't understand what it describes.)
The same makes it hard to determine whether a patent is even valid.
The rule about wilful infringement meaning triple damages would make it difficult to publish the results of such a study.
Microsoft, of course, could settle the question much more quickly, by just telling us which of their patents (they believe) Linux infringes. That's assuming they have any patents that would survive a court case. They must have, mustn't they? They wouldn't have said it if it wasn't true...
I just gave them $20. Paypal sent me to their project page, which has some Google adverts on the right-hand side. Guess whose site is at the top of that panel? That's right - kamind.com... the same guy who's saying copyright doesn't protect their software. I'm sure he'll appreciate helping to pay for their defence...
...unfortunately. No matter how cleverly you hide your address from the bots, the humans that you actually want to hear from have to enter the real thing into their email client. If the client stores the address in its address book, or it keeps a copy of the message, any piece of malware that infects the user's machine can discover your address and transmit it back to Spam Central for bombardment with the latest round of pump-n-dump.
I'm convinced this is how those bastards got the address of mine that currently gets the most spam. I maintain two sites, each with a contact address. They're minimally obfuscated - instead of user@example.com it says user at example dot com. One address gets almost no legitimate mail, but almost no spam. The other gets one legitimate mail every month or so, and spam maybe once a week. (Oh yes, I count myself lucky. My spam load peaked at a hundred a day a few years ago.)
I wonder if there would be any mileage in a mail client that encrypted the address book and mail folders, so that other processes running under your user ID couldn't read them? Trouble is, anyone savvy enough to choose a client because it has such a feature probably isn't going to get hit by malware in the first place. Good luck getting this feature into Outlook and switched on by default...
I agree that panspermia smacks of argument from incredulity - "We can't think of a mechanism for life to start on Earth, so we'll say it started somewhere else and hitched a ride on a meteor to get here." Not really much different from "God did it," or "The Flying Spaghetti Monster did it." It doesn't propose an answer for how the life that came from somewhere else started. Until we have a better idea of how widespread life is in the universe, and how similar any of it is to us, we can't say for sure that the theory is wrong.
But one reason that the theory is appealing to some is that the fossil record seems to show that life appeared on Earth not long after the planet became capable of supporting life. If life originated here, you have only a small amount of time and space in which to get all the chemistry right. Advocates of panspermia would have you believe that the odds of getting it right are low. But if you accept that life might have started elsewhere, that allows more space (other planets, or even dust clouds) and more time (stars older than the Sun) for that origin to occur. The odds of life starting at least once in that larger time and space are much better than the odds of it starting on one specific planet in one specific period. It only has to happen once. After that, by definition, life will spread.
I'm not saying I believe it myself; just trying to explain why some people do. And this simulation doesn't prove that the theory is right, just that it's possible.
That is not the patent application that TFA is talking about. Patent number 6,990,452 has the following people listed as inventors and is assigned to AT&T:
Ostermann; Joern (Morganville, NJ); Civanlar; Mehmet Reha (Middletown, NJ); Cosatto; Eric (Highlands, NJ); Graf; Hans Peter (Lincroft, NJ); LeCun; Yann Andre (Lincroft, NJ)
Whereas patent application number 20060015812 has only two inventors and is (or will be) assigned to Cingular Wireless:
Both of them are concerned with emoticons, but that's as far as the similarity goes. The '452 patent is about using them as hints in a text-to-speech system that has an animated face reading your messages to you, so that the computer can make the face smile or frown or whatever at the appropriate point. The '812 application is about a menu system for inserting emoticons into messages.
True. Bear in mind that this kind of system proves where your phone was - not necessarily where you were.
I know of at least one case where positioning information from a mobile has helped to clear someone, that of Damilola Taylor. Four youngsters were accused of murdering an 11-year-old boy. A mobile belonging to one of the defendants was used two miles from the scene of the crime, seven minutes beforehand, and it seemed there was no way he could have covered the distance quickly enough. (It later transpired that there was a shortcut that the prosecution didn't know about or didn't consider.)
The main reason the defendants were cleared was actually that a key prosecution witness was found to have lied about something. The judge decided that her testimony was unreliable and ordered the jury to acquit two of the defendants. They later found the other two not guilty.
In Soviet Russia, they didn't have God telling them how He'd designed the world and everything in it. Instead they had Comrade Lysenko telling them how to increase agricultural yields through methods that sounded plausible but didn't have a hope of working. It mightn't have been so bad if he didn't have the ears of Uncle Joe and the party machinery...
if the guy were to actually make something out of this then doesn't he deserve the patent?
The whole point is that the laws of physics, as we currently understand them, say that he can't make anything out of it, any more than he can make a perpetual motion machine.
If it's "complete malarky" then nobody has anything to worry about...
You can say that because, as a Slashdotter, you know how broken the patent system is. Joe Public, though, probably still thinks that getting a patent on something means it must be genuine and useful. I see three possible reasons for the "inventor" applying for this patent:
He wants to demonstrate the system's brokenness. Fine: we'll cheer him on from the sidelines.
He's a crackpot. Having the patent lets him say, "Look! My theory is right! Experts(*) agree that a device can be built that utilises it!" That is, his crackpottery now appears more respectable than before.
He's a conman, trying to lure investors into throwing money at him, with the promise of enormous profits once the device goes into production - which of course it never will. The patent lends an air of legitimacy to the scam.
(2 and 3 shade into one another somewhat.) The Patent Office should be able to realise when someone is making a joke at its expense. It should not be assisting people who want to perpetrate junk science or fraud.
(* note that he doesn't say what the experts' area of expertise is. I wouldn't like to say either, but it's certainly not physics.)
I should say that I probably don't have the whole picture. My expertise is in performance management, which is mainly concerned with ensuring that when someone wants to make a call, it connects successfully, and the network holds onto it until one party voluntarily hangs up. (My job really is to get that information in front of someone who knows what it means and what to do about it.) So when I say that some data isn't available, or isn't recorded, it might actually be in some other place that my customers aren't interested in.
If you lose the signal while you're making a call, the data that I deal with will contain a record of that. If the network decides to pass your call over to a nearby cell - especially because the signal is weak or poor quality - the data will record that as well.
Your phone monitors the strength and quality of the signal from nearby cells, to help the network decide which cell the phone should "camp" on. (This cell is the one it will first contact when you try to make a call.) This data would obviously help the company to find out where coverage is poor, but they don't routinely keep it - thousands of cells times thousands of measurements per phone per day times millions of phones means a lot of data to sift through.
If you switch your phone off, the phone notifies the network of that, but this is more to help the network know that it should divert to voicemail when someone calls you. The network can't assume that you're switching your phone off because coverage is bad: your battery might be low, or you might not want to be disturbed for a while. I suppose the phone could send a similar message to tell the network that it was about to go out of coverage, but I don't believe that this happens.
The phone company has to keep records about where each call originated so that they can bill you properly. The network automatically keeps aggregate statistics about what's happened in each cell - how many times someone tried to make a call, how many calls connected, how many calls completed successfully, how many were cut off because of signal loss, that sort of thing.
If they see a high rate of failure coming from one cell, they can tell the network to gather more detailed information about it for a while. The reason they don't do this automatically is because the data can add up to gigabytes per cell per day, and nearly all of it's useless. The difficulty isn't in parsing the data - I do that for a living - but in storing and analysing it quickly enough that the results aren't so old as to be irrelevant.
Fixing problems of signal strength or quality can be just a matter of reorienting an antenna by a few degrees, but putting up a new tower in an area that already has one is not a trivial task. The kit is expensive, and they cost a lot to run. Most companies would want to be reasonably sure that the new tower will pay for itself. Then they have to find somewhere to put it. Then they have to persuade whoever owns the land or the building to let them put it there. Then they have to argue the case with the local authorities. Then they have to win over residents who think the tower will give them cancer and lower their property values...
I'll grant you this is difficult, but I don't believe it's impossible. That's why I said "with (the appropriate versions of) the appropriate tools." It would normally be more trouble than it was worth, but in theory, you could tell the user the exact versions of all the tools and libraries that were involved in compilation, and the exact command lines to type, and that would give them a binary that was identical to yours. Usually, the source would be available for download, and I suppose anyone who cares about having it would probably compile it themselves without looking at any precompiled binaries that you supplied. But since you were talking about making only the binary available for download, and sending the source on a CD, the user has to ask themselves: "How do I know that this source is really that which coyote-san used to create this binary?" The only way they can be sure is to try to compile an identical binary - which they can't, because the sources aren't the same.
I also don't follow your argument that nobody can do the same thing with my code.
I was assuming that you own the copyright in the code. (It becomes more complicated if you accept patches from other people, so I'll assume you wrote the whole thing yourself for now.) As the owner, you can do things with the code that other people are forbidden from doing without your permission. For instance, if you put your binary on some server for people to download, you're under no obligation to provide the source for exactly that binary to anybody. If they ask for the source to that binary and you give them source that's the same except for the addition of the serial number, this is fine (unless you have a contract with them that says otherwise). Conceivably, they could sue you for false or misleading advertising, but this is nothing to do with the GPL.
Now, if you tell your users that the source code you're giving them is GPL'd, then any user who distributes a binary of that source has to make that exact source available to anyone who downloads their binary. They don't own the copyright, so they require the copyright owner's permission to distribute the program (binary or source). The GPL is what gives them that permission, and they have to abide by its terms if they want that permission to continue to exist. If you wanted to allow your users to distribute binaries and sources that didn't correspond with each other, then you (as copyright owner) could do this, but then you would not be licensing your program under the GPL. The license might look very similar to the GPL, but it would not actually be the GPL.
Having said all of that, several others in this thread have pointed out that successfully violating the GPL is harder than many people seem to think. Apart from all the clues that are left behind in the binary, there's always the possibility that someone who works for the infringing company will blow the whistle on them.
Then again, in a small market, would there be any benefit to having a hidden serial number? Company X requests a CD of your source, and you send them one. Three months later, company X launches an all-new product that looks suspiciously like yours. You buy a copy, run "strings" on the binaries, and find the same unique serial number that was in the source you sent them. Great. You sue the living daylights out of them and retire to a tropical island. But what do you do if their serial number isn't there? Do you conclude that there's no copyright infringement - that their product's similarity to yours is just the result of reverse engineering and the fact that they're chasing the same customers as you?
Or are you expecting that you might find company Y's serial number in company X's product? While Y might be devious enough to try to get X into trouble by giving them their CD of your source, I can't imagine that X would be stupid enough to go along with it, knowing
This is a nice idea, but strictly, I believe it would. The parts of the GPL that talk about making source code available use the words "corresponding source code." I would interpret this to mean "that source code which, when the user compiles it with (the appropriate versions of) the appropriate tools, produces binaries that are identical to those that the user already has." If the binary differs by so much as one bit, then it's not the "corresponding source code."
Even if your hidden serial number is completely benign, and it's the only difference in your source code, how is the user to know that? How is the user to decide whether the binary he compiled himself is "functionally identical" to the one you compiled for him? The general case of this problem is unsolvable. Incidentally, can you be sure that your serial number will actually appear in the binary? The compiler (or the user) might notice that it's not being used anywhere, and decide to omit it.
Then again... if it's your code, you can do whatever you want with it - including things that the GPL would forbid other people from doing. So you can tell everybody who downloads your software that it's GPL'd, but then when they ask for the source, give them a copy with a serial number. If anybody else was to do this with your code, that would be a GPL violation.
For several months, any time I found an "unsubscribe" link in a spam that took you to a web page where you type in the address you want to "unsubscribe"... I would type in one or more of:
It's in London, on the morning of 14 December. Would any UK or EU Slashdotters who can't attend like me to ask a question on their behalf? Post them in reply to this, and I'll see what I can do.
Publish your ideas in a reputable academic journal - one that somebody working in your field would be expected to read. You need to be able to prove, maybe years later, that your ideas were available to the public on a certain date.
If someone later manages to get a patent on your idea(s), you can file a request with the USPTO to ask them to strike it down or narrow its scope. You can cite your publication as prior art that the PTO should have considered. The Microsoft FAT patent was invalidated in just this way.
The relevant section of the law is 35 USC 301. You have to pay a fee to file the form, but it's pretty small. For added poetic justice, go to the small claims court and sue the bastard who patented your idea to get the fee back.
Slashdot Mirror
I thought they put that in as an Easter egg... the Patriot Act isn't a novel. Though some Eastern bloc countries allegedly used 1984 as a HOWTO, or a specification of an ideal government.
That's pretty much the definition of a vanity press. PublishAmerica (who were nearly duped into publishing Atlanta Nights) are just a bit better known than most.
No one will admit to running a vanity press, which is why they have to be exposed, and naive authors warned against them.
There are legitimate companies where you can pay to have your book published, but these guys are up-front about what they offer and how much it costs. They don't pretend to be choosy about what they print. (400 pages of /dev/urandom? No problem.) They also tend to be realistic about your chances of having a bestseller, or even turning a profit (comparable to the odds of finding Hamlet's soliloquy in those 400 pages).
I agree that the committee's ideas about what to fund seem strange sometimes, though the "strange" Good Causes will get more media attention than the "normal" ones. Most of the money paid into the lottery goes straight back out to those who buy tickets, in the form of prizes.
The money spent on tickets is distributed as follows:
(From http://www.camelotfoundation.org.uk/camelot.asp )
28% isn't exactly a very small percentage. Though who knows - maybe the Government's 12% should go to the Good Causes fund too. Maybe Branson could run it more efficiently than Camelot, but if the Government won't give up its share, a Branson-run lottery would give between 28% and 33% of the sales to good causes. That doesn't sound like so much of an improvement to me. And that's assuming ticket sales don't go down when he takes over.
I know the plural of "anecdote" is not "data", but Firefox 2 has been very stable for me. This is on Mandriva 2007. The only extension I use is NoScript.
Mozilla 1.x would consistently crash after a week or two, with an X error that made me think it had run out of some kind of resource.
I agree Firefox 2 is a bit of a memory hog. Mine is using about 350Mb at the moment, with 8 tabs open, though most of that is swapped out. Maybe I'm getting old, but it feels fast enough. It sounds as though version 3 will be a pleasant surprise.
I live in the UK, and our minister in charge of education really is called Mr Balls. I have to remind myself not to snigger every time I hear about him.
"Happy slapping" is the act of assaulting someone while one of your friends films it, usually on a camera phone, the aim being to send the recording to all your other friends to prove how tough you are (or something like that)...
That would probably be The Metamorphosis of Prime Intellect, by the same localroger chappie. (Too lazy to dig out a link to it.)
I read it a couple of years ago and, while the plot was interesting, I wouldn't say I enjoyed it very much. I'm not keen on blood and gore, and I think the large amount of it in the first half of the book put me off.
Probably it is an urban legend, but I thought that one was fairly sensible, given who's likely to be wearing the costume. Consider: you're four years old and like watching Superman on TV. You've figured out that Superman and Clark Kent are really the same person, but for some reason, Clark never outruns (outflies?) a speeding bullet or leaps tall buildings in a single bound. What's the difference? Clark just wears a boring blue suit like millions of other guys, but Superman wears this cool costume with the underpants outside. Obviously, therefore, the costume must be the source of Superman's powers... [*]
[*] I didn't believe this when I was four. I read the comics, where you can see that Clark has the Superman costume underneath his suit. Besides, you couldn't buy superhero costumes in those days, and I must have known that one I improvised wouldn't have the special powers of the real thing...
More often than I care to remember :-)
I can't really argue with that. If someone is attacking you, you'd be foolish not to defend yourself. (Or run away!) But surely it's better if you can prevent them from being allowed to attack you in the first place. Both approaches can proceed in parallel, of course. It just makes me angry that people such as those at WikiPatents have to spend so much time and effort cleaning up other people's incompetence and malice, when they could be making positive contributions to the useful arts and sciences.
I fear you're right there. I try to think with my head more than my gut. That's probably one reason I'm a software engineer, not a CIO :-) "Prove it or shut up" works when your adversary is the prosecution in a criminal trial, or a scientist advocating a new theory. But I suppose when your opponent is shovelling great steaming piles of FUD onto you or your (potential) customers, you have to do something to clear it away. I'd love to be a fly on the wall when the first CIO asks a Microsoft sales rep, "So, which of your patents does Linux infringe on, then?"
Good points... if you want to continue the metaphor, remember that Britain had an empire in 1914 too... the biggest the world had ever seen. (The saying "The sun never sets on the British Empire" meant literally that - whatever time of day or night it was in London, it would be daytime for some part of the Empire.) Nowadays, our possessions consist of Northern Ireland and a few islands that function mainly as tax havens and tourist traps.
I can't really see a big company starting a patent war, unless they think they have no alternative. Suppose Microsoft were to sue Red Hat, claiming that Red Hat Linux infringes one or more Microsoft patents. Before the ink was dry on the writs, IBM would be telling Microsoft to drop the suit or face a suit for infringement of (say) several hundred IBM patents. The thing is, Microsoft knows that would (almost certainly) happen, so they're not going to sue Red Hat unless they can be very sure of winning against IBM.
Here, as before World War I, you have an enormous buildup of armed forces on all sides. But I think the reason World War I escalated so quickly and became so devastating was alliances between countries and, worse, secret alliances. Two or more countries would promise to help one another if either was attacked... but they wouldn't tell anyone else about it. If several countries enter such treaties, that pretty much guarantees that any war will escalate until it involves everybody.
The laws governing companies make it difficult (in theory) for publicly-traded companies to form secret alliances. And patents are all visible to everybody, so someone who's thinking of starting a patent war can see what the opposition could decide to use against them.
I don't want a patent war, but it might be the only way to make lawmakers see how broken the patent system has become. It seems humanity never learns a lesson except the hard way...
I'm aware of the triple damages rule. It was the last bullet point in my list.
There's a wiki about patents, called, appropriately enough, WikiPatents. I don't know much more about them; I just found them linked from pubpat.org today. They're looking at all US patents, not just those related to software. What's on the site might not be admissible in a court case, but it would certainly give a head start to a lawyer who's trying to get a patent invalidated.
Worthy though such projects are, they're treating the symptoms, not the disease. Why the hell should the public have to do the patent office's job for it? The real solution is to stop bad patents from being issued in the first place.
But really, as this comment points out, since Microsoft have made the allegation, it's up to them to prove it. The free software community doesn't have to disprove it. We don't have to do anything beyond saying (repeatedly), "Tell us which features of which projects infringe which of your patents. Then we will get those patents invalidated, or work around them, or remove the infringing code, in that order. Until you give us the information we have asked for, we will have to assume that either there is no infringement, or you do not want the infringement cured. Both alternatives imply you are acting in bad faith."
The article author is conflating patents and copyright. Is it too much to ask that someone who (presumably) gets paid to write this stuff would know the difference? From the summary:
Having access to Microsoft's source code would neither help nor hinder anyone in deciding whether Linux infringes any Microsoft patents. The whole point of patents is that they're visible to the public. The US government even maintains a database of them that anyone with a web browser can look at! A patent holder doesn't have to implement the invention that their patent describes before they can sue someone else for infringing it. If they did, patent trolls couldn't exist.
Anyone could drive a stake through the heart of Microsoft's FUD campaign by getting a list of all patents held by Microsoft and proving, for each patent, either (a) Linux does not infringe the patent or (b) the patent is invalid. I suspect this will not happen, for any or all of the following reasons:
Microsoft, of course, could settle the question much more quickly, by just telling us which of their patents (they believe) Linux infringes. That's assuming they have any patents that would survive a court case. They must have, mustn't they? They wouldn't have said it if it wasn't true...
I just gave them $20. Paypal sent me to their project page, which has some Google adverts on the right-hand side. Guess whose site is at the top of that panel? That's right - kamind.com... the same guy who's saying copyright doesn't protect their software. I'm sure he'll appreciate helping to pay for their defence...
...unfortunately. No matter how cleverly you hide your address from the bots, the humans that you actually want to hear from have to enter the real thing into their email client. If the client stores the address in its address book, or it keeps a copy of the message, any piece of malware that infects the user's machine can discover your address and transmit it back to Spam Central for bombardment with the latest round of pump-n-dump.
I'm convinced this is how those bastards got the address of mine that currently gets the most spam. I maintain two sites, each with a contact address. They're minimally obfuscated - instead of user@example.com it says user at example dot com. One address gets almost no legitimate mail, but almost no spam. The other gets one legitimate mail every month or so, and spam maybe once a week. (Oh yes, I count myself lucky. My spam load peaked at a hundred a day a few years ago.)
I wonder if there would be any mileage in a mail client that encrypted the address book and mail folders, so that other processes running under your user ID couldn't read them? Trouble is, anyone savvy enough to choose a client because it has such a feature probably isn't going to get hit by malware in the first place. Good luck getting this feature into Outlook and switched on by default...
I agree that panspermia smacks of argument from incredulity - "We can't think of a mechanism for life to start on Earth, so we'll say it started somewhere else and hitched a ride on a meteor to get here." Not really much different from "God did it," or "The Flying Spaghetti Monster did it." It doesn't propose an answer for how the life that came from somewhere else started. Until we have a better idea of how widespread life is in the universe, and how similar any of it is to us, we can't say for sure that the theory is wrong.
But one reason that the theory is appealing to some is that the fossil record seems to show that life appeared on Earth not long after the planet became capable of supporting life. If life originated here, you have only a small amount of time and space in which to get all the chemistry right. Advocates of panspermia would have you believe that the odds of getting it right are low. But if you accept that life might have started elsewhere, that allows more space (other planets, or even dust clouds) and more time (stars older than the Sun) for that origin to occur. The odds of life starting at least once in that larger time and space are much better than the odds of it starting on one specific planet in one specific period. It only has to happen once. After that, by definition, life will spread.
I'm not saying I believe it myself; just trying to explain why some people do. And this simulation doesn't prove that the theory is right, just that it's possible.
That is not the patent application that TFA is talking about. Patent number 6,990,452 has the following people listed as inventors and is assigned to AT&T:
Ostermann; Joern (Morganville, NJ); Civanlar; Mehmet Reha (Middletown, NJ); Cosatto; Eric (Highlands, NJ); Graf; Hans Peter (Lincroft, NJ); LeCun; Yann Andre (Lincroft, NJ)
Whereas patent application number 20060015812 has only two inventors and is (or will be) assigned to Cingular Wireless:
Cunningham; Ivy; (Seattle, WA) ; White; Christopher; (Redmond, WA)
Both of them are concerned with emoticons, but that's as far as the similarity goes. The '452 patent is about using them as hints in a text-to-speech system that has an animated face reading your messages to you, so that the computer can make the face smile or frown or whatever at the appropriate point. The '812 application is about a menu system for inserting emoticons into messages.
True. Bear in mind that this kind of system proves where your phone was - not necessarily where you were.
I know of at least one case where positioning information from a mobile has helped to clear someone, that of Damilola Taylor. Four youngsters were accused of murdering an 11-year-old boy. A mobile belonging to one of the defendants was used two miles from the scene of the crime, seven minutes beforehand, and it seemed there was no way he could have covered the distance quickly enough. (It later transpired that there was a shortcut that the prosecution didn't know about or didn't consider.)
The main reason the defendants were cleared was actually that a key prosecution witness was found to have lied about something. The judge decided that her testimony was unreliable and ordered the jury to acquit two of the defendants. They later found the other two not guilty.
In Soviet Russia, they didn't have God telling them how He'd designed the world and everything in it. Instead they had Comrade Lysenko telling them how to increase agricultural yields through methods that sounded plausible but didn't have a hope of working. It mightn't have been so bad if he didn't have the ears of Uncle Joe and the party machinery...
The whole point is that the laws of physics, as we currently understand them, say that he can't make anything out of it, any more than he can make a perpetual motion machine.
You can say that because, as a Slashdotter, you know how broken the patent system is. Joe Public, though, probably still thinks that getting a patent on something means it must be genuine and useful. I see three possible reasons for the "inventor" applying for this patent:
(2 and 3 shade into one another somewhat.) The Patent Office should be able to realise when someone is making a joke at its expense. It should not be assisting people who want to perpetrate junk science or fraud.
(* note that he doesn't say what the experts' area of expertise is. I wouldn't like to say either, but it's certainly not physics.)
I should say that I probably don't have the whole picture. My expertise is in performance management, which is mainly concerned with ensuring that when someone wants to make a call, it connects successfully, and the network holds onto it until one party voluntarily hangs up. (My job really is to get that information in front of someone who knows what it means and what to do about it.) So when I say that some data isn't available, or isn't recorded, it might actually be in some other place that my customers aren't interested in.
If you lose the signal while you're making a call, the data that I deal with will contain a record of that. If the network decides to pass your call over to a nearby cell - especially because the signal is weak or poor quality - the data will record that as well.
Your phone monitors the strength and quality of the signal from nearby cells, to help the network decide which cell the phone should "camp" on. (This cell is the one it will first contact when you try to make a call.) This data would obviously help the company to find out where coverage is poor, but they don't routinely keep it - thousands of cells times thousands of measurements per phone per day times millions of phones means a lot of data to sift through.
If you switch your phone off, the phone notifies the network of that, but this is more to help the network know that it should divert to voicemail when someone calls you. The network can't assume that you're switching your phone off because coverage is bad: your battery might be low, or you might not want to be disturbed for a while. I suppose the phone could send a similar message to tell the network that it was about to go out of coverage, but I don't believe that this happens.
The phone company has to keep records about where each call originated so that they can bill you properly. The network automatically keeps aggregate statistics about what's happened in each cell - how many times someone tried to make a call, how many calls connected, how many calls completed successfully, how many were cut off because of signal loss, that sort of thing.
If they see a high rate of failure coming from one cell, they can tell the network to gather more detailed information about it for a while. The reason they don't do this automatically is because the data can add up to gigabytes per cell per day, and nearly all of it's useless. The difficulty isn't in parsing the data - I do that for a living - but in storing and analysing it quickly enough that the results aren't so old as to be irrelevant.
Fixing problems of signal strength or quality can be just a matter of reorienting an antenna by a few degrees, but putting up a new tower in an area that already has one is not a trivial task. The kit is expensive, and they cost a lot to run. Most companies would want to be reasonably sure that the new tower will pay for itself. Then they have to find somewhere to put it. Then they have to persuade whoever owns the land or the building to let them put it there. Then they have to argue the case with the local authorities. Then they have to win over residents who think the tower will give them cancer and lower their property values...
I'll grant you this is difficult, but I don't believe it's impossible. That's why I said "with (the appropriate versions of) the appropriate tools." It would normally be more trouble than it was worth, but in theory, you could tell the user the exact versions of all the tools and libraries that were involved in compilation, and the exact command lines to type, and that would give them a binary that was identical to yours. Usually, the source would be available for download, and I suppose anyone who cares about having it would probably compile it themselves without looking at any precompiled binaries that you supplied. But since you were talking about making only the binary available for download, and sending the source on a CD, the user has to ask themselves: "How do I know that this source is really that which coyote-san used to create this binary?" The only way they can be sure is to try to compile an identical binary - which they can't, because the sources aren't the same.
I was assuming that you own the copyright in the code. (It becomes more complicated if you accept patches from other people, so I'll assume you wrote the whole thing yourself for now.) As the owner, you can do things with the code that other people are forbidden from doing without your permission. For instance, if you put your binary on some server for people to download, you're under no obligation to provide the source for exactly that binary to anybody. If they ask for the source to that binary and you give them source that's the same except for the addition of the serial number, this is fine (unless you have a contract with them that says otherwise). Conceivably, they could sue you for false or misleading advertising, but this is nothing to do with the GPL.
Now, if you tell your users that the source code you're giving them is GPL'd, then any user who distributes a binary of that source has to make that exact source available to anyone who downloads their binary. They don't own the copyright, so they require the copyright owner's permission to distribute the program (binary or source). The GPL is what gives them that permission, and they have to abide by its terms if they want that permission to continue to exist. If you wanted to allow your users to distribute binaries and sources that didn't correspond with each other, then you (as copyright owner) could do this, but then you would not be licensing your program under the GPL. The license might look very similar to the GPL, but it would not actually be the GPL.
Having said all of that, several others in this thread have pointed out that successfully violating the GPL is harder than many people seem to think. Apart from all the clues that are left behind in the binary, there's always the possibility that someone who works for the infringing company will blow the whistle on them.
Then again, in a small market, would there be any benefit to having a hidden serial number? Company X requests a CD of your source, and you send them one. Three months later, company X launches an all-new product that looks suspiciously like yours. You buy a copy, run "strings" on the binaries, and find the same unique serial number that was in the source you sent them. Great. You sue the living daylights out of them and retire to a tropical island. But what do you do if their serial number isn't there? Do you conclude that there's no copyright infringement - that their product's similarity to yours is just the result of reverse engineering and the fact that they're chasing the same customers as you?
Or are you expecting that you might find company Y's serial number in company X's product? While Y might be devious enough to try to get X into trouble by giving them their CD of your source, I can't imagine that X would be stupid enough to go along with it, knowing
This is a nice idea, but strictly, I believe it would. The parts of the GPL that talk about making source code available use the words "corresponding source code." I would interpret this to mean "that source code which, when the user compiles it with (the appropriate versions of) the appropriate tools, produces binaries that are identical to those that the user already has." If the binary differs by so much as one bit, then it's not the "corresponding source code."
Even if your hidden serial number is completely benign, and it's the only difference in your source code, how is the user to know that? How is the user to decide whether the binary he compiled himself is "functionally identical" to the one you compiled for him? The general case of this problem is unsolvable. Incidentally, can you be sure that your serial number will actually appear in the binary? The compiler (or the user) might notice that it's not being used anywhere, and decide to omit it.
Then again... if it's your code, you can do whatever you want with it - including things that the GPL would forbid other people from doing. So you can tell everybody who downloads your software that it's GPL'd, but then when they ask for the source, give them a copy with a serial number. If anybody else was to do this with your code, that would be a GPL violation.
For several months, any time I found an "unsubscribe" link in a spam that took you to a web page where you type in the address you want to "unsubscribe"... I would type in one or more of:
I also imagine Darl (or his secretary, more likely) has been getting rather a lot of phone calls offering to refinance his mortgage. Does that help?
It's in London, on the morning of 14 December. Would any UK or EU Slashdotters who can't attend like me to ask a question on their behalf? Post them in reply to this, and I'll see what I can do.
Publish your ideas in a reputable academic journal - one that somebody working in your field would be expected to read. You need to be able to prove, maybe years later, that your ideas were available to the public on a certain date.
If someone later manages to get a patent on your idea(s), you can file a request with the USPTO to ask them to strike it down or narrow its scope. You can cite your publication as prior art that the PTO should have considered. The Microsoft FAT patent was invalidated in just this way.
The relevant section of the law is 35 USC 301. You have to pay a fee to file the form, but it's pretty small. For added poetic justice, go to the small claims court and sue the bastard who patented your idea to get the fee back.