Well, almost every ISP out there already does this. In the US, if you choose to run anything other than Windows and have any problems the usual response from an ISP is to go "We don't support Windows." and refuse to help you until you boot into Windows to troubleshoot. And if the problem doesn't occur in Windows, it's simply not a problem as far as they're concerned. Lundis Energi is simply doing exactly the same thing, except that this time it's Windows users on the other end of "We don't support that.". Sauce for the goose and all.
On all media? Maybe. Is a downloadable file a different media from a real-time stream? If it is and YouTube wants to support browsers that don't have streaming plug-ins but do have viewers, they'll need that right.
In all formats? Definitely. WMA is a different format than QuickTime is a different format than RealMedia, and if they want to offer translation of the content into a format the user can view without requiring the uploader to supply all possible formats they're going to need that right.
Sublicenseable? If YouTube France is a different business unit from YouTube USA and you want content uploaded to the US site to be viewable on the French site (a reasonable assumption, they're both YouTube after all), then they'll need to be able to grant the French business unit the neccesary rights that you've granted the US unit.
The problem is that YouTube does need all those rights, but it doesn't need them for unlimited purposes. It's just really really hard to phrase the limitations in a way that both makes sense and allows enough leeway for everyone involved to do what the uploader wants done. Normally this is all handled by being a bit fuzzy about it: you both agree that this whole deal's supposed to be for a particular purpose and everything's supposed to be read to be implicitly limited to being for that purpose. The result is language that's entirely reasonable, but that looks really really scary when taken out of the context that provides the limitation. My only concern would be to make sure that the agreement did in fact state the purpose and intent, so I have something in writing later in case I need it.
Yes, but think about this: you uploaded your video to YouTube specifically so it could be shown to others. But showing it to others constitutes a performance of that video by YouTube. Now, if they don't include in the terms something saying you give them the right to perform your video, how are they going to show it to others? They've no right to performance, you didn't give them one, so they can't do the very thing you want them to do with your video. Similarly, if you don't give them the right to present it in a different form, they can't include it on their front page or provide people the ability to embed the video (see any number of blog pages where, instead of a link to YouTube, the blogger embeds the actual video in a playable form). And if you don't give YouTube the right to make copies of your video, they can't make the multiple copies onto their cache and delivery servers at various points on the network close to the destination networks (think how Akamai works).
One can argue the exact wording (I prefer terms that make it explicit that the grant is for the sole purpose of providing content within the Web site and related operations and that uses outside the context of the Web site aren't part of the grant), but copyright law means that YouTube and the like have to ask you for certain rights simply to be able to legally do what you want them to do.
Not quite. It means a vendor doing that and allowing itself the ability to update and change the software while at the same time preventing the user from doing the same. If the vendor simply made it impossible to change the software, for either the user or the vendor, then that'd be acceptable under the GPL. The GPLv3 is explicit about this: it's not a violation to put the software in ROM or the like that can't be changed, but if the software can be changed then the recipient must be able to change it. TiVo's issue is that they want the right to change GPL'd software themselves but not permit the people they distribute it to to do what TiVo does.
I don't think anyone's ever taken a case on on that point. They're all interested in trying to argue that the EULA's not valid, or that their agreement to it should be set aside. I don't recall anyone ever arguing that the EULA's irrelevant because they never agreed to it and don't have to agree to it.
Were I cynical, I'd suggest that this might be because it'd turn it from a sexy First Amendment case into a bog-standard Uniform Commercial Code contract case that wouldn't make for nearly as good a headline for the lawyers.
#2 won't happen. Remember that I'm not filtering anything. I'm simply declaring (in named.conf for my network's nameserver) that I control say "2o7.net" and that any query for an A record in that domain should return one resolving to address 192.168.171.68. The web server at that address returns a 404 Not Found error for any request. No filter in the browser, no proxy in the way, the only thing 2o7 can do is change their domain to something else (which I'll add to named.conf in turn). And it's fire-and-forget: I make one entry, kick Bind, and forget about the entire matter.
As for advertisers not wanting intrusive ads, if they don't then why are so many of them using more and more intrusive ones? Ads used to be small graphics. Now animated full-page-width banners are the least intrusive ones, and float-overs, pop-ups and worse are common. One Web comic I know is currently in a fight with their ad network: they specified no pop-ups but the network's serving up banners that, when moused over, expand to fill the entire browser window and won't contract until the mouse moves off of the now-larger ad. This sounds to me like exactly the opposite of what you're claiming, and I have to wonder about this disconnect between your claims and observed reality.
That "marginal fringe" includes just about everybody I talk to. I don't know anybody (except other Unix geeks like me) who doesn't gripe about the problems associated with ads and/or doesn't ask me if I'd give them some help cleaning up their computer. They usually don't know where the problems came from, and they all have one question: "Can you do anything to keep it from ever happening again?". I'm seriously looking into how much it'd cost to make a suitable mini-box for people.
I was a lot less strange and brittle before having to clean the 50th Gator/Claria, 180Solutions/Zango, HuntBar, CoolWebSearch or similar infestation, along with all the other crud those things had installed after the behind-the-back installation from a Web advertisement. I don't have the time or energy to be dealing with that on a daily basis. So I can either engage in a prolonged slowly-escalating battle of filters vs. advertiser's tactics, or I can nuke the problem from orbit and get on with more productive things. Option 2 takes less of my time, and the problems it creates for advertisers... are Not My Problem.
if you're not going to click on an ad, there's no reason to use ad block either... why block something that is not forced on you? it seems to be an issue for a fringe group of strange brittle characters
Actually, my reason for going beyond just using an ad-blocker and completely usurping the ad-server's domain in DNS is that just seeing the ad is the problem. Anymore, ads go beyond just an image on a page. They're nasty, distracting animations. They're pop-up windows that clutter the screen. They're those floating "windows" that lay on top of the content that you came to the page to see. And we won't even get into the amount of malware and crudware they try to install (I routinely get them trying to install ActiveX controls without cluing anybody in that they're installing them, and I do not tolerate random Web sites installing executables on my machines).
I could use a simple ad-blocker, I suppose, but it's a losing fight with advertisers getting nastier and nastier about things. So I decided that if it's just going to escalate anyway, I might as well go straight to 11: a Web server that returns 404 for any request, a local-network DNS server that declares itself authoritative for the entire domain of any advertiser I notice, and some IP redirection for advertisers sneaky enough to use IP addresses rather than DNS names (or to use names in the hosting site's domain pointing at their servers). Nasty to set up initially, but it's cut down the amount of crud I have to deal with on Windows boxes by an order of magnitude.
My objection to the law, and to Arnold's appeal of the ruling, is that the law doesn't do anything about the problem it purports to address. Sure, it bans the sale of certain games to minors. As a practical matter, most stores won't sell those games to minors anyway. And in every case that's come up as the motivation for these laws, the stores didn't sell the games to a minor. They sold the game to a legal adult, most often a parent of the minor involved, and that adult then gave the game to the minor. Now, what's the store supposed to do when a legal adult comes in to buy those games? They aren't a minor, the store doesn't know what they're going to do with the game after they buy it, on what grounds is the store supposed to refuse to sell it to them?
And I notice the people pushing these laws aren't also pushing to punish the parents who bought the games for their kids. If the stores aren't supposed to sell those games to minors, why are adults who give the games to minors exempt?
That depends. If MySQL AB has accepted any contributions to the MySQL code and didn't get copyright assignment on them, then MySQL AB is bound by the GPL itself. No, they aren't bound by section 3a, but that just means they're bound by 3b (since 3c isn't available to them and section 3 describes the only GPL-acceptable ways to distribute).
b) Accompany it with a written offer, valid for at least three years, to give any third party, for a charge no more than your cost of physically performing source distribution, a complete machine-readable copy of the corresponding source code, to be distributed under the terms of Sections 1 and 2 above on a medium customarily used for software interchange
Now, does that term say anything about the third party only being a customer of the distributor? No, it doesn't. It doesn't modify "any third party" in any way, so that phrase means what it says: any third party. If MySQL AB wants to distribute source only to it's own customers, it needs to distribute under section 3a (3c being unavailable to it since it's doing commercial distribution).
You forget the other place clause "c" comes into play. That's when a customer of MySQL AB receives binaries from them under clause "b" and then noncommercially under clause "c" passes along the written offer MySQL AB made to them to a recipient. That recipient is then entitled to the source code from MySQL AB, because they've got that written offer. The FSF considered your scenario, and wrote the GPL so that it doesn't work your way. There's a path from the final recipient up to the original source, and the party responsible for providing the final recipient with the source code is always on that path.
Note that this is really a special case of a more general case the GPL was worded to deal with. The general case is any distributor handing out binaries and pointing recipients at an uninvolved third party (eg. SourceForge) for the sources. That third party may not be under any obligation to keep the sources available (eg. they only distribute under clause "a") and so may decide to take the sources down. The GPL anticipated this, and insured that recipients can't be left with no way to get the source in that case.
Not technically correct. They can limit giving the source code to only their customers if and only if they provide the source code along with the binaries. If they provide the source code seperately, then the GPL requires them to offer the source code to any third party that asks for it for at least 3 years from their last binary distribution. This is because any party who receives the binary is entitled to the source even if they didn't get it directly from MySQL AB.
If they provide the source code along with the binaries, the GPL considers that to have satisfied their obligations. After that, they're not obliged to give the source code to anybody else. Not even customers.
Now, if they don't provide the source code with the binaries, if customers are obliged to get it separately from the binary package, then the obligation is to provide the source to anybody who asks for it, customer or not, and that obligation lasts for 3 years after the last binary was distributed. Note that if the binaries are available via download, offering the source for download at the same time and from the same page satisfies the GPL's requirement to provide source along with the binaries even if the customer doesn't actually download the source code at the time.
Answer: it's not. MS is just now picking up on what you can do with the Virtual 8086 mode introduced on the 80386 and used to good advantage in OS/2. The idea goes back even further, IBM's been doing this with the Virtual Machine Monitor on their mainframes since forever. Of course, one of the prerequisites to doing it efficiently is that applications have to avoid mucking with the hardware directly and only use system or library calls to do things, otherwise you have to fully virtualize the hardware instead of just providing call emulation within the VM. This bumps up against things like DirectX which are designed around just the sort of low-level hardware-specific access you want to avoid.
True, it's not a guarantee for any one person. But ask any insurance-company actuary and they'll tell you it's a statistical guarantee: in a group of a certain size, it's virtualy 100% certain that a given percentage of them will die in any given year. The whole basis of life insurance is being able to predict that percentage accurately enough to guarantee that your premium income will be more than your payouts due to deaths, and insurance companys have those numbers down to a science. They can't tell you who'll die among their policyholders, but they can tell you down to a fraction of a percent how many of their policyholders will die that year. If they couldn't, they couldn't make a profit and they're definitely making a profit.
The whole point of insurance is to spread the costs around. Not risks, costs. Let's take an example. Suppose there's a group of 1000 people, and in any given year one of them's going to get hit with a $100,000 bill. None of them can afford that large hit, but all of them can afford to pay $100 per year. So they start a pool, each paying in their $100 with the understanding that the pool will cover the entire bill for whichever of them gets unlucky that year. Sure, the other 999 have to pay even if they don't get hit that year, but they also avoid the even higher expense of preparing to handle that big bill and the worrying over what'll happen if they get unlucky before they've saved up enough to handle it.
Now, suppose the guy running the pool for everybody decides there's an awful lot of money floating around in the pool. He could, he thinks, work out which person'll be the unlucky one that year. If he can, then he can charge that person the full $100,000 that year. That'll cover the pay-out and leave the other $99,900 in the pool for him to play with. Yes, this is the extreme case, but it's what the insurance companies here want to do taken to it's logical conclusion.
But wait a minute. If I'm a member of the pool, the whole reason I'm paying my $100 every year is so I won't get hit with the high bill if my number happens to come up that year. If I'm going to get hit with that huge bill anyway, why am I paying in? I'm not getting any protection from it, I'd be better off with that extra $100 every year to spend myself. The more it moves towards that extreme case, the less reason I have to pay into the pool. And even at the near end, the more people decide to pull out of the pool the more the guy running it has to charge those who're left, which makes it less attractive for them to remain in the pool, which means more people will pull out. And when there's nobody left, who will the guy running the pool get his money from? Oops.
No. The idea behind insurance is to spread the risk out across a large group so that no individual member gets hit with a sudden cost they can't afford.
In a way, it's like a mortgage, but spread across people instead of time. Most people can't afford to shell out a quarter of a million dollars in one shot for a house, but they can afford to pay say $1200/month. So they take out a loan that costs them $1200/month for 30 years, get the house, and spread the cost across that entire 30 years. Same for insurance. Say you have a group of 1000 people. In any given year, you know that 1 of them will get hit with a $100,000 bill. None of them can afford that, it'd drive them into bankruptcy. But all of them can afford $100/year. So each year all of them pay $100 each into a pool, giving the pool $100,000 for the year, with the understanding that the pool will cover the bill for the one of them that gets unlucky that year. If it sounds like a bad deal for those that don't get unlucky, think about what they'd have to set aside each year to prepare to be able to handle the full bill on their own, and the worrying they'd have to do over what happens if they get unlucky before they've had enough time to prepare.
But now the guy running the pool's noticed that there's a lot of money in the pool. If he can figure out how to finger the guy who'll be unlucky and charge him the full $100,000 that year, he can cover what the pool has to pay out and still have $99,900 left from all the other payments that he can use for himself. That sounds like a really good deal for the guy running the pool. But wait. As a member of the pool, I'm paying my $100/year to know that I won't have to pay that $100,000 if and when. If I know I'll be stuck paying it anyway, then why should I bother joining the pool? It's costing me money without giving me any protection against that huge bill anymore. Since I'll have to face the possibility of that unaffordable hit either way, I'd be smarter to take my $100/year and spend it on something immediate instead of contributing to the pool. So everybody pulls out of the pool, and there goes the pool. Oops, there goes our pool-operator's big chunk of money.
You also better know what game you're supposed to be playing. This is Microsoft's big problem. The best football game in the world won't help you at a tennis match.
"That's a really great, uncounterable move... for chess. Pity we're playing checkers. *tak* *tak* *tak* King me."
One factor is the time window. We only invented radio about 130 years ago. To any civilization more than 130 light-years away, we're invisible in the radio bands since our first transmissions haven't reached them yet. Our transmission footprint's also interesting. Our output increased steadily up to a point, but more recently it's been decreasing as we move to more efficient transmission methods (more directional signals, tighter directional signals, non-electrical transmission methods like fiber-optics that don't generate RF). By the time we hit the 2- or 3-century mark (measured from when we started transmitting RF) it's likely we'll be emitting so little that we won't be visible to anyone who doesn't know exactly where to look and exactly what to look for. On top of that, we've only been listening for other civilizations in the radio band for about 50 years (a little less, actually). That gives us a 350 year "window". For a civilization N light-years away from us, they have to have invented radio between N and N+350 years ago for us to see them. If they invented it more recently than N years ago, their first signals won't have gotten to us yet. If they invented it more than N+350 years ago, the trailing edge of their detectable transmissions will have passed Earth before we started listening. The same works for us being visible to them: to any civilization more than 130 light-years away we don't exist in the radio bands because our first transmissions haven't reached them yet.
As far as anyone visiting us, I'd say that any civilization that's got feasible interstellar travel going isn't using radio or anything else primitive enough for us to detect anymore. And our solar system is a big place. To see any visitors we have to be looking at exactly the right spot at exactly the time they're there, and we aren't looking at a big percentage of the sky at any given time so it's easy to simply not be looking at the right spot or be looking at it at the wrong time. And look at our reactions to any evidence that might turn up. If I presented a broken plate of metal with a slightly odd composition as evidence of a visit from ET, the instant reaction of 99% of the planet would be to laugh me off as just another lunatic. And were I a visitor from an extra-solar civilization, I'd make sure that I didn't leave anything more than the occasional small chunk of debris or the occasional sighting too fleeting for anyone to get clear pictures from it. Those natives may be primitive, but fission warheads can still be really annoying so it's better to just stay discrete and avoid scaring the natives into doing something rash. Not to mention that if you're seriously observing a culture you probably want to not interfere with it lest you screw up your own data, so any cultural observers would likely be taking great pains to avoid being noticed at all.
I wouldn't be worried about not seeing extra-solar civilizations. My worries will start the day we do start seeing solid evidence of them visiting. Why? Well, based on the record on this planet, those emmissaries are more likely to be representatives of the ET equivalent of the British East India Company than anything else, and that's not likely to be good news for us.
My preferred choice is a thumb-operated trackball. Your thumb does get tired initially, it's not used to being used that way. It's no worse than any change in pointing device, though, and in my case after a week I'd adapted and was more comfortable with the trackball than a mouse. I've never been comfortable with conventional trackballs with the ball under the fingers.
I originally swore by the Logitech Trackman Wheel, but then Logitech changed the design to make it too narrow for my fingers to be comfortable on the buttons. I switched to a Microsoft Trackball Optical, 5 buttons and much wider and more comfortable, but then MS discontinued that model (I've got an extra stashed away, though). Lately I've switched one computer to a Logitech laser mouse because I just can't find a trackball that suits that's still in production.
Unfortunately you can't bill them. If you read the licenses for those BSA-member products, they contain a clause saying you agree they can audit you at any time, with no advance warning, at your expense. Doesn't matter whether they find any violations or not, you foot the bill.
Because it isn't IE7 that's being exploited. It's the part of Windows that matches URIs to programs to open them via registry entries. IE7 comes into it because those routines in Windows are really part of IE (remember that IE's an integral part of Windows). When you install IE7, you install a new system library with new implementations of those routines that replace the ones from IE6, and said new implementations contain the bug that's being exploited.
This is also a good illustration of why making core parts of your OS part of the browser is a bad idea.
Slashdot Mirror
Well, almost every ISP out there already does this. In the US, if you choose to run anything other than Windows and have any problems the usual response from an ISP is to go "We don't support Windows." and refuse to help you until you boot into Windows to troubleshoot. And if the problem doesn't occur in Windows, it's simply not a problem as far as they're concerned. Lundis Energi is simply doing exactly the same thing, except that this time it's Windows users on the other end of "We don't support that.". Sauce for the goose and all.
On all media? Maybe. Is a downloadable file a different media from a real-time stream? If it is and YouTube wants to support browsers that don't have streaming plug-ins but do have viewers, they'll need that right.
In all formats? Definitely. WMA is a different format than QuickTime is a different format than RealMedia, and if they want to offer translation of the content into a format the user can view without requiring the uploader to supply all possible formats they're going to need that right.
Sublicenseable? If YouTube France is a different business unit from YouTube USA and you want content uploaded to the US site to be viewable on the French site (a reasonable assumption, they're both YouTube after all), then they'll need to be able to grant the French business unit the neccesary rights that you've granted the US unit.
The problem is that YouTube does need all those rights, but it doesn't need them for unlimited purposes. It's just really really hard to phrase the limitations in a way that both makes sense and allows enough leeway for everyone involved to do what the uploader wants done. Normally this is all handled by being a bit fuzzy about it: you both agree that this whole deal's supposed to be for a particular purpose and everything's supposed to be read to be implicitly limited to being for that purpose. The result is language that's entirely reasonable, but that looks really really scary when taken out of the context that provides the limitation. My only concern would be to make sure that the agreement did in fact state the purpose and intent, so I have something in writing later in case I need it.
Yes, but think about this: you uploaded your video to YouTube specifically so it could be shown to others. But showing it to others constitutes a performance of that video by YouTube. Now, if they don't include in the terms something saying you give them the right to perform your video, how are they going to show it to others? They've no right to performance, you didn't give them one, so they can't do the very thing you want them to do with your video. Similarly, if you don't give them the right to present it in a different form, they can't include it on their front page or provide people the ability to embed the video (see any number of blog pages where, instead of a link to YouTube, the blogger embeds the actual video in a playable form). And if you don't give YouTube the right to make copies of your video, they can't make the multiple copies onto their cache and delivery servers at various points on the network close to the destination networks (think how Akamai works).
One can argue the exact wording (I prefer terms that make it explicit that the grant is for the sole purpose of providing content within the Web site and related operations and that uses outside the context of the Web site aren't part of the grant), but copyright law means that YouTube and the like have to ask you for certain rights simply to be able to legally do what you want them to do.
Two words to Polaris: procmail, formail.
Not quite. It means a vendor doing that and allowing itself the ability to update and change the software while at the same time preventing the user from doing the same. If the vendor simply made it impossible to change the software, for either the user or the vendor, then that'd be acceptable under the GPL. The GPLv3 is explicit about this: it's not a violation to put the software in ROM or the like that can't be changed, but if the software can be changed then the recipient must be able to change it. TiVo's issue is that they want the right to change GPL'd software themselves but not permit the people they distribute it to to do what TiVo does.
I don't think anyone's ever taken a case on on that point. They're all interested in trying to argue that the EULA's not valid, or that their agreement to it should be set aside. I don't recall anyone ever arguing that the EULA's irrelevant because they never agreed to it and don't have to agree to it.
Were I cynical, I'd suggest that this might be because it'd turn it from a sexy First Amendment case into a bog-standard Uniform Commercial Code contract case that wouldn't make for nearly as good a headline for the lawyers.
#2 won't happen. Remember that I'm not filtering anything. I'm simply declaring (in named.conf for my network's nameserver) that I control say "2o7.net" and that any query for an A record in that domain should return one resolving to address 192.168.171.68. The web server at that address returns a 404 Not Found error for any request. No filter in the browser, no proxy in the way, the only thing 2o7 can do is change their domain to something else (which I'll add to named.conf in turn). And it's fire-and-forget: I make one entry, kick Bind, and forget about the entire matter.
As for advertisers not wanting intrusive ads, if they don't then why are so many of them using more and more intrusive ones? Ads used to be small graphics. Now animated full-page-width banners are the least intrusive ones, and float-overs, pop-ups and worse are common. One Web comic I know is currently in a fight with their ad network: they specified no pop-ups but the network's serving up banners that, when moused over, expand to fill the entire browser window and won't contract until the mouse moves off of the now-larger ad. This sounds to me like exactly the opposite of what you're claiming, and I have to wonder about this disconnect between your claims and observed reality.
That "marginal fringe" includes just about everybody I talk to. I don't know anybody (except other Unix geeks like me) who doesn't gripe about the problems associated with ads and/or doesn't ask me if I'd give them some help cleaning up their computer. They usually don't know where the problems came from, and they all have one question: "Can you do anything to keep it from ever happening again?". I'm seriously looking into how much it'd cost to make a suitable mini-box for people.
I was a lot less strange and brittle before having to clean the 50th Gator/Claria, 180Solutions/Zango, HuntBar, CoolWebSearch or similar infestation, along with all the other crud those things had installed after the behind-the-back installation from a Web advertisement. I don't have the time or energy to be dealing with that on a daily basis. So I can either engage in a prolonged slowly-escalating battle of filters vs. advertiser's tactics, or I can nuke the problem from orbit and get on with more productive things. Option 2 takes less of my time, and the problems it creates for advertisers... are Not My Problem.
if you're not going to click on an ad, there's no reason to use ad block either... why block something that is not forced on you? it seems to be an issue for a fringe group of strange brittle characters
Actually, my reason for going beyond just using an ad-blocker and completely usurping the ad-server's domain in DNS is that just seeing the ad is the problem. Anymore, ads go beyond just an image on a page. They're nasty, distracting animations. They're pop-up windows that clutter the screen. They're those floating "windows" that lay on top of the content that you came to the page to see. And we won't even get into the amount of malware and crudware they try to install (I routinely get them trying to install ActiveX controls without cluing anybody in that they're installing them, and I do not tolerate random Web sites installing executables on my machines).
I could use a simple ad-blocker, I suppose, but it's a losing fight with advertisers getting nastier and nastier about things. So I decided that if it's just going to escalate anyway, I might as well go straight to 11: a Web server that returns 404 for any request, a local-network DNS server that declares itself authoritative for the entire domain of any advertiser I notice, and some IP redirection for advertisers sneaky enough to use IP addresses rather than DNS names (or to use names in the hosting site's domain pointing at their servers). Nasty to set up initially, but it's cut down the amount of crud I have to deal with on Windows boxes by an order of magnitude.
My objection to the law, and to Arnold's appeal of the ruling, is that the law doesn't do anything about the problem it purports to address. Sure, it bans the sale of certain games to minors. As a practical matter, most stores won't sell those games to minors anyway. And in every case that's come up as the motivation for these laws, the stores didn't sell the games to a minor. They sold the game to a legal adult, most often a parent of the minor involved, and that adult then gave the game to the minor. Now, what's the store supposed to do when a legal adult comes in to buy those games? They aren't a minor, the store doesn't know what they're going to do with the game after they buy it, on what grounds is the store supposed to refuse to sell it to them?
And I notice the people pushing these laws aren't also pushing to punish the parents who bought the games for their kids. If the stores aren't supposed to sell those games to minors, why are adults who give the games to minors exempt?
That depends. If MySQL AB has accepted any contributions to the MySQL code and didn't get copyright assignment on them, then MySQL AB is bound by the GPL itself. No, they aren't bound by section 3a, but that just means they're bound by 3b (since 3c isn't available to them and section 3 describes the only GPL-acceptable ways to distribute).
To quote the GPLv2, section 3b:
b) Accompany it with a written offer, valid for at least three years, to give any third party, for a charge no more than your cost of physically performing source distribution, a complete machine-readable copy of the corresponding source code, to be distributed under the terms of Sections 1 and 2 above on a medium customarily used for software interchange
Now, does that term say anything about the third party only being a customer of the distributor? No, it doesn't. It doesn't modify "any third party" in any way, so that phrase means what it says: any third party. If MySQL AB wants to distribute source only to it's own customers, it needs to distribute under section 3a (3c being unavailable to it since it's doing commercial distribution).
GPLv3 has similar language, it's in section 6b.
You forget the other place clause "c" comes into play. That's when a customer of MySQL AB receives binaries from them under clause "b" and then noncommercially under clause "c" passes along the written offer MySQL AB made to them to a recipient. That recipient is then entitled to the source code from MySQL AB, because they've got that written offer. The FSF considered your scenario, and wrote the GPL so that it doesn't work your way. There's a path from the final recipient up to the original source, and the party responsible for providing the final recipient with the source code is always on that path.
Note that this is really a special case of a more general case the GPL was worded to deal with. The general case is any distributor handing out binaries and pointing recipients at an uninvolved third party (eg. SourceForge) for the sources. That third party may not be under any obligation to keep the sources available (eg. they only distribute under clause "a") and so may decide to take the sources down. The GPL anticipated this, and insured that recipients can't be left with no way to get the source in that case.
Not technically correct. They can limit giving the source code to only their customers if and only if they provide the source code along with the binaries. If they provide the source code seperately, then the GPL requires them to offer the source code to any third party that asks for it for at least 3 years from their last binary distribution. This is because any party who receives the binary is entitled to the source even if they didn't get it directly from MySQL AB.
If they provide the source code along with the binaries, the GPL considers that to have satisfied their obligations. After that, they're not obliged to give the source code to anybody else. Not even customers.
Now, if they don't provide the source code with the binaries, if customers are obliged to get it separately from the binary package, then the obligation is to provide the source to anybody who asks for it, customer or not, and that obligation lasts for 3 years after the last binary was distributed. Note that if the binaries are available via download, offering the source for download at the same time and from the same page satisfies the GPL's requirement to provide source along with the binaries even if the customer doesn't actually download the source code at the time.
Answer: it's not. MS is just now picking up on what you can do with the Virtual 8086 mode introduced on the 80386 and used to good advantage in OS/2. The idea goes back even further, IBM's been doing this with the Virtual Machine Monitor on their mainframes since forever. Of course, one of the prerequisites to doing it efficiently is that applications have to avoid mucking with the hardware directly and only use system or library calls to do things, otherwise you have to fully virtualize the hardware instead of just providing call emulation within the VM. This bumps up against things like DirectX which are designed around just the sort of low-level hardware-specific access you want to avoid.
True, it's not a guarantee for any one person. But ask any insurance-company actuary and they'll tell you it's a statistical guarantee: in a group of a certain size, it's virtualy 100% certain that a given percentage of them will die in any given year. The whole basis of life insurance is being able to predict that percentage accurately enough to guarantee that your premium income will be more than your payouts due to deaths, and insurance companys have those numbers down to a science. They can't tell you who'll die among their policyholders, but they can tell you down to a fraction of a percent how many of their policyholders will die that year. If they couldn't, they couldn't make a profit and they're definitely making a profit.
The whole point of insurance is to spread the costs around. Not risks, costs. Let's take an example. Suppose there's a group of 1000 people, and in any given year one of them's going to get hit with a $100,000 bill. None of them can afford that large hit, but all of them can afford to pay $100 per year. So they start a pool, each paying in their $100 with the understanding that the pool will cover the entire bill for whichever of them gets unlucky that year. Sure, the other 999 have to pay even if they don't get hit that year, but they also avoid the even higher expense of preparing to handle that big bill and the worrying over what'll happen if they get unlucky before they've saved up enough to handle it.
Now, suppose the guy running the pool for everybody decides there's an awful lot of money floating around in the pool. He could, he thinks, work out which person'll be the unlucky one that year. If he can, then he can charge that person the full $100,000 that year. That'll cover the pay-out and leave the other $99,900 in the pool for him to play with. Yes, this is the extreme case, but it's what the insurance companies here want to do taken to it's logical conclusion.
But wait a minute. If I'm a member of the pool, the whole reason I'm paying my $100 every year is so I won't get hit with the high bill if my number happens to come up that year. If I'm going to get hit with that huge bill anyway, why am I paying in? I'm not getting any protection from it, I'd be better off with that extra $100 every year to spend myself. The more it moves towards that extreme case, the less reason I have to pay into the pool. And even at the near end, the more people decide to pull out of the pool the more the guy running it has to charge those who're left, which makes it less attractive for them to remain in the pool, which means more people will pull out. And when there's nobody left, who will the guy running the pool get his money from? Oops.
No. The idea behind insurance is to spread the risk out across a large group so that no individual member gets hit with a sudden cost they can't afford.
In a way, it's like a mortgage, but spread across people instead of time. Most people can't afford to shell out a quarter of a million dollars in one shot for a house, but they can afford to pay say $1200/month. So they take out a loan that costs them $1200/month for 30 years, get the house, and spread the cost across that entire 30 years. Same for insurance. Say you have a group of 1000 people. In any given year, you know that 1 of them will get hit with a $100,000 bill. None of them can afford that, it'd drive them into bankruptcy. But all of them can afford $100/year. So each year all of them pay $100 each into a pool, giving the pool $100,000 for the year, with the understanding that the pool will cover the bill for the one of them that gets unlucky that year. If it sounds like a bad deal for those that don't get unlucky, think about what they'd have to set aside each year to prepare to be able to handle the full bill on their own, and the worrying they'd have to do over what happens if they get unlucky before they've had enough time to prepare.
But now the guy running the pool's noticed that there's a lot of money in the pool. If he can figure out how to finger the guy who'll be unlucky and charge him the full $100,000 that year, he can cover what the pool has to pay out and still have $99,900 left from all the other payments that he can use for himself. That sounds like a really good deal for the guy running the pool. But wait. As a member of the pool, I'm paying my $100/year to know that I won't have to pay that $100,000 if and when. If I know I'll be stuck paying it anyway, then why should I bother joining the pool? It's costing me money without giving me any protection against that huge bill anymore. Since I'll have to face the possibility of that unaffordable hit either way, I'd be smarter to take my $100/year and spend it on something immediate instead of contributing to the pool. So everybody pulls out of the pool, and there goes the pool. Oops, there goes our pool-operator's big chunk of money.
You also better know what game you're supposed to be playing. This is Microsoft's big problem. The best football game in the world won't help you at a tennis match.
"That's a really great, uncounterable move... for chess. Pity we're playing checkers. *tak* *tak* *tak* King me."
One factor is the time window. We only invented radio about 130 years ago. To any civilization more than 130 light-years away, we're invisible in the radio bands since our first transmissions haven't reached them yet. Our transmission footprint's also interesting. Our output increased steadily up to a point, but more recently it's been decreasing as we move to more efficient transmission methods (more directional signals, tighter directional signals, non-electrical transmission methods like fiber-optics that don't generate RF). By the time we hit the 2- or 3-century mark (measured from when we started transmitting RF) it's likely we'll be emitting so little that we won't be visible to anyone who doesn't know exactly where to look and exactly what to look for. On top of that, we've only been listening for other civilizations in the radio band for about 50 years (a little less, actually). That gives us a 350 year "window". For a civilization N light-years away from us, they have to have invented radio between N and N+350 years ago for us to see them. If they invented it more recently than N years ago, their first signals won't have gotten to us yet. If they invented it more than N+350 years ago, the trailing edge of their detectable transmissions will have passed Earth before we started listening. The same works for us being visible to them: to any civilization more than 130 light-years away we don't exist in the radio bands because our first transmissions haven't reached them yet.
As far as anyone visiting us, I'd say that any civilization that's got feasible interstellar travel going isn't using radio or anything else primitive enough for us to detect anymore. And our solar system is a big place. To see any visitors we have to be looking at exactly the right spot at exactly the time they're there, and we aren't looking at a big percentage of the sky at any given time so it's easy to simply not be looking at the right spot or be looking at it at the wrong time. And look at our reactions to any evidence that might turn up. If I presented a broken plate of metal with a slightly odd composition as evidence of a visit from ET, the instant reaction of 99% of the planet would be to laugh me off as just another lunatic. And were I a visitor from an extra-solar civilization, I'd make sure that I didn't leave anything more than the occasional small chunk of debris or the occasional sighting too fleeting for anyone to get clear pictures from it. Those natives may be primitive, but fission warheads can still be really annoying so it's better to just stay discrete and avoid scaring the natives into doing something rash. Not to mention that if you're seriously observing a culture you probably want to not interfere with it lest you screw up your own data, so any cultural observers would likely be taking great pains to avoid being noticed at all.
I wouldn't be worried about not seeing extra-solar civilizations. My worries will start the day we do start seeing solid evidence of them visiting. Why? Well, based on the record on this planet, those emmissaries are more likely to be representatives of the ET equivalent of the British East India Company than anything else, and that's not likely to be good news for us.
My preferred choice is a thumb-operated trackball. Your thumb does get tired initially, it's not used to being used that way. It's no worse than any change in pointing device, though, and in my case after a week I'd adapted and was more comfortable with the trackball than a mouse. I've never been comfortable with conventional trackballs with the ball under the fingers.
I originally swore by the Logitech Trackman Wheel, but then Logitech changed the design to make it too narrow for my fingers to be comfortable on the buttons. I switched to a Microsoft Trackball Optical, 5 buttons and much wider and more comfortable, but then MS discontinued that model (I've got an extra stashed away, though). Lately I've switched one computer to a Logitech laser mouse because I just can't find a trackball that suits that's still in production.
Unfortunately you can't bill them. If you read the licenses for those BSA-member products, they contain a clause saying you agree they can audit you at any time, with no advance warning, at your expense. Doesn't matter whether they find any violations or not, you foot the bill.
Because it isn't IE7 that's being exploited. It's the part of Windows that matches URIs to programs to open them via registry entries. IE7 comes into it because those routines in Windows are really part of IE (remember that IE's an integral part of Windows). When you install IE7, you install a new system library with new implementations of those routines that replace the ones from IE6, and said new implementations contain the bug that's being exploited.
This is also a good illustration of why making core parts of your OS part of the browser is a bad idea.