What would be the equivalent local rate for scanners in Europe?
Probably about $35 an hour, they'd only work seven hours, three days a week, and they'd be on strike half the year anyway. And you can't fire any of them.;-)
Actually, as this is being done in association with university libraries, I think they shouldn't have any problems getting reliable help at $10USD an hour, because that's significantly more than a lot of other on-campus jobs pay. I know from personal experience that many of the students that get paid to videotape campus events and have access to thousands of dollars of semi-pro videography gear are only paid $8-10 an hour. Same for stage electricians, scene shop carpenters and painters, and audio technicians. You'd be pretty amazed at the level of responsibility that people are given, despite being paid less than a typical mouth-breathing fast food restaurant zombie.
If they're going to hire college students, I think the problem is more about finding people who are interested in the work and not in it just to sit and listen to their iPod while making some beer bucks. That means that the pay should be consistent with the skill level relative to other campus jobs, but not much more than that. (Or at least don't advertise it; use the salary as a method of retention, not recruitment.)
IMHO, it seems like a little cooperation here would make a lot of sense for both parties - they could save money trading digital copies 1-for-1 while remaining in (healthy) competition.
This is very true. However you see this sort of thing in a lot of emerging industries -- two competitors will duplicate each other's work until eventually one defeats the other in the marketplace and buys up their work at fire-sale prices. As long as either one thinks that they can "win," there's little incentive to help.
Too bad though, because the other thing they could do is divide the 'contentspace' and let Google deal with the stuff that's under copyright (which has a lot of thorny legal issues but is potentially a gold mine in terms of rewards) while letting the OCA do the stuff that's free, and link their databases through one Search interface. Two backends, one frontend. They could even pull Project Gutenberg's collection and the Internet Archive in there too.
But I think that as long as Yahoo is championing one effort and Google the other, there won't be that much integration between them.
Commercial or non-commercial use doesn't enter into it.
If the work in question is under copyright, you can't copy and redistribute it; if it's not, then you can. The only exceptions would be the fair use provisions, and I don't think that they would cover you reproducing an entire book, even if it was for non-commercial use: if you're a university professor you can't copy an entire textbook and give them out to your students. That's a non-commerical use, but it's still illegal. There might be some exceptions for purely personal use -- some type of "format shifting" perhaps, like OCRing it and running it through text-to-speech and putting the result on your iPod -- that you could make a good case for if you already owned the book in print form, but non-commerical use normally isn't an excuse for infringement. Despite public opinion to the contrary, there is no exception to a copyright holder's exclusive rights for "non-commercial" uses.
Also, if you made a derivative work from something that was out of copyright, and then went and tried to sell it, only the portions that you contributed anew would be protected, the existing stuff doesn't change. That's not to say that you couldn't sell it (if it's out of copyright you don't even have to change anything to sell it, you can go and print out anything you want from Project Gutenberg and try to sell it, if you think you'll get any buyers), but you wouldn't have any recourse against someone taking your changed version, editing out all of your changes, and selling it themselves.
There is a fairly good introduction to these concepts here. Or read it straight from the U.S. Code.
If you bought a copy of any classic book that is out of copyright, and it's a literal republication of the original (not a 'modern interpretation' or new translation or anything else) than you could, I believe, scan, OCR, and distribute the resulting text. The literary work -- be it Shakespeare's, Clemens', Dickens', etc. -- is no longer protected by copyright.
You could not, however, scan the book and distribute the images of the pages. Because although the original author's text is not under copyright protection, the book itself (layout, design, etc.) could be. Also, any changes they might have made to the text (new grammar, or diction) could be, which is why you'd have to be careful. I think it would only qualify as a new protected work if the changes represented "an original work of authorship" according to 17 U.S.C. 101, but depending on the publisher they might try to sue you into bankruptcy anyway.
As long as you didn't copy the layout or any of the additional materials (critical essays, introductions) that publishers put into re-prints of classic literature, I see no reason why it would be illegal to type in and share the $2.99 Penguin Classics edition of Tom Sawyer that you can get at any Borders.
No, that's the point of a NAT router with a firewall.
A device doing pure NAT would forward all traffic -- including unsolicited incoming traffic -- on the public IP of the router to your computer's internal IP. So yes, you're right, someone on the outside wouldn't know your computer's internal IP address, but it wouldn't matter because the NAT box would happily forward all the packets. Because that's what NAT means.
Most home router boxes have NAT with selective port forwarding, plus a stateful-firewall-esque feature where it doesn't forward packets that aren't part of a connection that was initiated from inside the network. These are strictly speaking not part of NAT, they just tend to be rolled into the same devices making people think of them as one unit. (On most routers you can turn these things off and get "pure NAT" by putting your computer in as the DMZ address, this gets you NAT without firewalling so you can run a server.) But all of them could be easily implemented on IPv6 with a firewall.
The perceived security of NAT comes from the firewall-type features that are built into most home routers, not from the address translation service itself. There's no security gained simply by not letting an outside attacker know your internal network address if your gateway passes all packets, and conversely none given away by letting your attacker know it, IF you have a good firewall that's set to reject unsolicited traffic.
With IPv6, users would still probably want to have a box sitting between their easily-owned Windows box and the public net, but instead of being a NAT+Firewall, it would just be a Firewall. To them, the use would be exactly the same, except that they would have real end-to-end connectivity when they wanted it, and still retain the same level of security (or lack thereof) that they have today.
I think people are reaching too far, envisioning whole-house DC lines and stuff that wasn't a great idea when Thomas Edison was pushing for it, and it's not any better today. You don't need low-voltage wiring running all over your house, because most of those low-voltage devices are clustered together. Lots of them in the computer room, maybe a bunch in the entertainment/TV room, maybe a few more in the bedroom. Not too many in the garage, very few in the workshop/laundry/utility room. Plus there are lots of things which still need AC (HVAC blower, fridge, oven, clothes dryer, anything that makes heat, almost all lighting) and you'd be wasting a lot of copper running twice the lines. It's overkill to run wires throughout the whole house when a few strategically placed rectifers could do the job.
However it is equally ridiculous to have each device make its own DC, if they all want the same voltage (or close to it). Personally I think 24 or 48VDC would be a good standard, because then you have headroom enough to use a DC-DC converter to step down or regulate the power to whatever the device needs. For computer peripherals I can't think of anything that wouldn't be able to run off of that voltage except for CRT displays, and they're going away anyway.
I would be all for a standardized, DC power connector for computer and entertainment equipment and peripherals. What I'd like to see is some sort of connector where the physical size of the connector is proportional to the current draw from the equipment. This is something I once saw on an old theatre lighting system: the receptacles were actually two buss bars, spaced an inch or so apart; the plugs were rectangular blocks with metal on the two long opposing sides, so that when you plugged them into the gap between the buss bars you'd connect to them. But the neat feature was that the plugs were longer if they drew more current: a 20A circuit would be say 4" long, and a 10A one only 2". If the house mains was only capable of supplying 200A, then there would only be 40" of space to plug into. It makes it very hard to overload the mains. They are also a bit scary to work with, since it's possible to stick fingers/objects/etc into the gap between the buss bars and create a short circuit across the entire incoming mains (which might come unfused right from the power company's transformer), and I think this is why they're not used anymore. OT: If anyone knows what this type of connector was called, I'd be interested to know.
Anyway, if I was going to design a DC power plug standard it would be with some system like that (but less apt to kill you), maybe with the recifier output being through two very narrow parallel slots, with some sort of a flexible insulating barrier keeping objects out. But the key thing would be that a high-current device (like a LCD monitor or CPU) would use a bigger plug than a peripheral, and the PS would only have enough room for as much current as it could produce. (And yes, you'd have to make 'plug splitters' illegal under a system like this.)
Converting between high and low voltages with DC power is much more difficult, and requires more complex equipment. (An AC transformer is two pieces of wire wrapped around a chunk of iron.)
It's my understanding too that even a good DC/DC converter is significantly less efficient than a transformer, so that the advantage you'd get by transmitting high voltage DC instead of AC would be lost when it came time to step the voltage back down to something useable.
DC is something you want to make at the point where there are a bunch of things that all want the same voltage. If all the equipment in your datacenter wants 12 and 24VDC, then go ahead and have one big (or several redundant) power supply for the 'center and distribute the DC. But if there are a lot of things which want different voltages you're better off sticking with AC and doing the voltage conversion before you rectify it.
I could easily see having a big power supply for a whole rack of CPUs, if all of them are the same type of motherboard and take the same voltages (+/-5, 12, etc.) but then all you've done is essentially DIYed a really space-inefficient blade server.
The other question is an economy of scale one: I'm not sure that the demand for large DC power supplies will be large enough, in the near future, to make them cheaper per watt than small (less than 1000W) ATX ones, especially if you want redundancy, at least up until you start talking about very large supplies like the ones for IBM z/Series mainframes and the like.
It was on the afternoon 'news in review' (Afternoon Edition?) a few days ago as well, plus there's been meta-coverage of the rootkit in stories about the amount of negative press Sony's received as a result of the rootkit and about the CA and NY lawsuits. I wouldn't be surprised if the stories about the lawsuits would have gotten coverage in business oriented publications (WSJ and others) although I don't subscribe to them so I can't check for sure.
From what I understand of the DRM scheme, once you install those kernel extensions, it keeps you from ripping the CD from there on out.
But the obvious flaw is that most users are going to rip the CD before they get around to listening to or checking out the additional content -- I know quite a few people who have iPods and their Macs are set up so that when you put in an Audio CD, iTunes is automatically launched, and when iTunes detects a new music CD, it automatically rips it into their Library. So by the time they'd get around to looking at that Data volume with the DRM installer, the music is already on their iPod.
I think the end result of this setup is that people will accidentally install the DRM after the fact, and never notice until for some reason they want to rip it again (maybe years later, if they haven't done a clean install of their system) and not be able to -- and I think to most people their first response will be to assume that the CD is scratched or otherwise bad, not that the computer is (since it rips all other music fine, and even ripped the CD fine originally).
Yeah I was about to say the same thing, you beat me to it. I think it's more likely that the kids will have the admin password and thus keep their clueless parents from messing up the computer. Guess it depends on the age of the kids and parents involved. Maybe younger parents with more computer experience with young kids would actually be in charge, but there are still a lot of families out there where the parents involvement in computer maintenance is limited to handing over their Visa card once in a while.
Anyone want to get a petition going for a Linux version?
Or is someone going to have to reverse-engineer the binary and produce a OpenRootkit that can automatically compile and execute itself on all POSIX-compliant systems?
If the application is just an executable that stands alone, generally there is not a permission dialog. You can copy that to anyplace on your file system, no problem at all. The exception is if you are a non-administrator user and you try to copy it to the Applications folder (which by default non-admins do not have write privs to).
Where you see the 'enter password' dialog consistently is when installing software using an actual Installer package -- usually you will not be allowed to go past the first step in the "Wizard" type thing (which just describes what the package is) without authenticating. In general, the Installer packages are only necessary when an application requires files to be put in various places around the system, especially in the Library, and isn't just a standalone application bundle. I have definitely seen programs which could have just been distributed as an application on a disk image using an Installer, but they're the exception to the rule (in this case, laziness on the developer's part produces the desired result, since most people won't make an Installer package if they don't have to -- why some people do, I don't know).
I've seen this but only on the Selectric I, and only on certain models -- the keyboard has the same number of keys as the later versions I believe, but the caps are labeled differently and the top row goes "~ 2 3 4 5 6 7 8 9" and then some special keys which vary. The one I've seen that was set up this way had some special keys on the one to the right of the numeral nine, for typing Library of Congress index card special characters that I've never seen on anything else. (A special type-ball was required also.)
I've never been able to figure out if the eight-number top row was standard on all of the very first generation of Selectrics, or was something they only did on special models, in order to squeeze special characters onto the keyboard that otherwise would have been left off.
No, other way around. Even if the EULA says that you can't resell the software, at least in some jurisdictions in the U.S. you can go ahead and legally do it anyway, because of what's known as the doctrine of "first sale."
Microsoft would probably grumble about it, but there's not a lot they could do, since the courts have ruled.
Basically I think the point of the original post was "So, what?" Meaning that this isn't some huge groundbreaking deal -- all that's new is that some company got MS's permission to resell software licenses, with the implication that MS is getting a cut, and people are acting like it's some big blow to the Evil Machine, when in reality people elsewhere could sell used software already, with or without permission.
That at least was my understanding of the implications of the comment.
My understanding was that IPv6 plays nicely with DHCP-like systems, because of its hierarchical nature. So your ISP would give you an address where the last portion was the MAC address of your router/gateway, and it would assign addresses out to your computer/fridge/toaster/etc. that had those units' MAC addresses in them. So the numbers are assigned from where a NAT gateway used to be, but they're globally (in theory) routable.
I'm not getting that information from the official spec or anything, it's more the informal discussions I've had with people who think it's a good idea, so feel free to take it with a grain of salt.
You can be sure that if I thought there was going to be any sort of creepy central control or registration involved, especially with the U.N. or something like it, that I'd be the last person on the planet to adopt IPv6, regardless of the technical advantages. However that's the first I've heard of that line of thinking so I'm not about to start oiling up the assault rifle and breaking out the Spam quite yet.;)
I don't think that people do this over gasoline (well, somewhere, somebody probably does) but they sure do it over other products. There's the never-ending, epic battle between Ford and Chevy enthusiasts, for one. I've met photographers who were as zealous about Nikon camera equipment (and eager to convert everyone else) as any Mac or Windows user. (Not to mention Kodak film vs. Fuji...) Videographers go back and forth on whether Panasonic cameras are a better deal than Sony, or if the latter are worth the price. In most gun clubs or stores you can get a spirited argument going by suggesting that Smith and Wesson handguns are superior to Colt's. At a cooking school you could probably get your ear talked off as to whether German or French chefs knives are better, and within those which brands are best. I could go on and on.
The quick answer to your question is 'yes.' Whenever you get people who spend a large percentage of their life in one industry, they develop preferences that seem obsessively odd to outsiders. It is our own fixation on computers that makes us think that people aren't just like this about other things; but being a "geek" isn't restricted to computers, we just don't use that term for people whose interest goes towards other things.
I'm not sure that they'd have to, because the addresses aren't given out randomly, they're hierarchical.
So really they'd only have to try a few million addresses that are sub-addresses of your ISP, at least this is based on my understanding of how the protocol is going to work.
But still you have a point, it would be like trying to find someone's Ethernet MAC address by exhausting that entire solution space (at least, since the IPv6 address will contain the MAC one).
It's not much of a cartel -- you will always have companies like Speakeasy, who are willing to toss in a few extras in order to siphon some customers off of Comcast or the local DSL telco and up to a higher price point.
As long as we don't let one company acquire and control all of the "last mile" bandwidth distribution networks, and we keep the cost of entrance into the ISP market reasonably low, you won't ever have that tight a cartel. Especially once you get the market fully penetrated and the flow of new customers starts to taper off, companies will start to think about how they can take them from each other, and that's when you'll see the features become a selling point.
I think the threat of a cartel is somewhat overstated, as long as we're vigilant about keeping the big telcos out of bed with the regulatory apparatus. (Or getting them out of bed, as the case may be.)
there's no need for NAT, and so there will be no NAT for IPv6
You obviously have never seen some geeks around here. There is no need for IP over carrier pigeon, and yet there is.
NAT is common enough an idea and -- despite the fact that I find it kludgey -- people seem to like it enough that I have no doubt that people will implement it on IPv6. It's not like it would be incredibly difficult or anything, especially if you dedicated a computer to doing it (as opposed to demanding it in some small-format dedicated appliance). Sure, maybe you'll never be able to go out to Best Buy and get an IPv6 NAT box for $30, because there won't be that level of demand for it, but for somebody willing to roll their own I'm sure it'll exist.
This assumes that there's no competition between ISPs. Eventually you'll get some value-added DSL or wireless internet provider that will give you that feature, and in time if there's personal demand for it -- which I think there will be, because as interactive content becomes more pervasive people will want the end-to-end communication ability that you can't get with NAT -- it will spread.
I think the reason why ISPs don't offer static IPs right now also has to do with an assumption that people who get static IPs will run their own servers, which tend to be high traffic and put a lot of load on the network. So that "static IP fee" is really being used to subsidize traffic-handling, it's not just paying for that address (the cost to the ISP of which is essentially zero).
Although I'm not saying that ISPs don't gouge the consumer pretty hard, I don't think that they have enough of a cartel to sink a feature like personal static IPs, if it was technically feasible at little to no cost.
But the more appropriate analogy is: You don't take your car in for complete engine rebuild if the engine is running fine.
While this may be true for your car, it's definitely not true of a helicopter, or a generator at a power plant, or any other important piece of machinery.
Would you still fly on an airline if that was their attitude towards maintenance? "Nah, we're not going to tear down that turbine...it hasn't failed yet!"
I think perhaps you should reevaluate the importance of the Internet to our society today. I think we've well surpassed the relative importance of a car to an average driver.
Neither of these points are really arguments for the current system, if anything they're good arguments against it, and in favor of IPv6.
#1 is nothing but a direct consecquence of the current shortage of IPv4 addresses. Under IPv6, there'd be no reason why every device on your network couldn't get a separate "real" address. The way they're handed out -- using a hierarchy instead of finite blocks -- would allow your ISP to let your home DHCP router hand out globally addressable IPs if it was set up correctly. Assuming your ISP doesn't suck, that is, and that's really not the fault of the IP system, one way or the other.
#2 is pretty frightening, because it shows a misunderstanding of what NAT is and a certain amount of laziness about security in general. That said, there's no reason why you couldn't get a 'firewall in a box' that would provide just as much (or as little) security without the NAT facility. It's just that right now when you go and buy a "home firewall" from Linksys, it almost always includes NAT by default (because of point #1, the pressure by ISPs on home users to only have one IP address due to limited supply). There's no reason why this needs to be true, however, and the security comes from the firewall effect and not the address translation itself.
I beg to differ. I question whether you're serious or a troll, but I'll respond anyway and give you the benefit of the doubt.
Lots of companies which are big enough to have their own Class-A allocations assign all of their clients globally routable addresses. I can tell you this from personal experience.
They don't use personal firewalls, obviously, and I have no idea why you think this is related. Using a personal firewall at the client level has nothing to do with IP address allocation or NAT. You can assign every user on a subnet a globally unique IP address, and then still use a stateful firewall for security. This is what these companies do: you get the benefit of not having your applications negotiate NAT with the protection of firewalls separating the internal networks at various facilities from the global network.
As far as the cost thing, if you're big enough to have a Class A block, you're not paying individually for IP addresses, so there's no difference in cost between a client that has a unique address and a NAT one. In fact the NAT one is probably slightly more expensive because the NAT routers are probably more maintainance and support-intensive than a straight firewall.
In short, I don't think you know what you're talking about. You might be correct when it comes to small or medium businesses, who are buying their connection from an ISP who is going to charge them more for a lot of static IPs than a few dynamic ones that they can use with NAT, but this issue isn't relevant to IBM, Ford, Apple, or the rest of the Class A companies.
Slashdot Mirror
Was I the only one who appreciated the irony when she said "This book almost killed me," and then went out to smoke a cigarette?
... she did mention that her previous job involved stapling sticks of gum to flyers for a club, so I guess she's moving up in the world.
Oh, well
What would be the equivalent local rate for scanners in Europe?
;-)
Probably about $35 an hour, they'd only work seven hours, three days a week, and they'd be on strike half the year anyway. And you can't fire any of them.
Actually, as this is being done in association with university libraries, I think they shouldn't have any problems getting reliable help at $10USD an hour, because that's significantly more than a lot of other on-campus jobs pay. I know from personal experience that many of the students that get paid to videotape campus events and have access to thousands of dollars of semi-pro videography gear are only paid $8-10 an hour. Same for stage electricians, scene shop carpenters and painters, and audio technicians. You'd be pretty amazed at the level of responsibility that people are given, despite being paid less than a typical mouth-breathing fast food restaurant zombie.
If they're going to hire college students, I think the problem is more about finding people who are interested in the work and not in it just to sit and listen to their iPod while making some beer bucks. That means that the pay should be consistent with the skill level relative to other campus jobs, but not much more than that. (Or at least don't advertise it; use the salary as a method of retention, not recruitment.)
IMHO, it seems like a little cooperation here would make a lot of sense for both parties - they could save money trading digital copies 1-for-1 while remaining in (healthy) competition.
This is very true. However you see this sort of thing in a lot of emerging industries -- two competitors will duplicate each other's work until eventually one defeats the other in the marketplace and buys up their work at fire-sale prices. As long as either one thinks that they can "win," there's little incentive to help.
Too bad though, because the other thing they could do is divide the 'contentspace' and let Google deal with the stuff that's under copyright (which has a lot of thorny legal issues but is potentially a gold mine in terms of rewards) while letting the OCA do the stuff that's free, and link their databases through one Search interface. Two backends, one frontend. They could even pull Project Gutenberg's collection and the Internet Archive in there too.
But I think that as long as Yahoo is championing one effort and Google the other, there won't be that much integration between them.
Commercial or non-commercial use doesn't enter into it.
If the work in question is under copyright, you can't copy and redistribute it; if it's not, then you can. The only exceptions would be the fair use provisions, and I don't think that they would cover you reproducing an entire book, even if it was for non-commercial use: if you're a university professor you can't copy an entire textbook and give them out to your students. That's a non-commerical use, but it's still illegal. There might be some exceptions for purely personal use -- some type of "format shifting" perhaps, like OCRing it and running it through text-to-speech and putting the result on your iPod -- that you could make a good case for if you already owned the book in print form, but non-commerical use normally isn't an excuse for infringement. Despite public opinion to the contrary, there is no exception to a copyright holder's exclusive rights for "non-commercial" uses.
Also, if you made a derivative work from something that was out of copyright, and then went and tried to sell it, only the portions that you contributed anew would be protected, the existing stuff doesn't change. That's not to say that you couldn't sell it (if it's out of copyright you don't even have to change anything to sell it, you can go and print out anything you want from Project Gutenberg and try to sell it, if you think you'll get any buyers), but you wouldn't have any recourse against someone taking your changed version, editing out all of your changes, and selling it themselves.
There is a fairly good introduction to these concepts here. Or read it straight from the U.S. Code.
If you bought a copy of any classic book that is out of copyright, and it's a literal republication of the original (not a 'modern interpretation' or new translation or anything else) than you could, I believe, scan, OCR, and distribute the resulting text. The literary work -- be it Shakespeare's, Clemens', Dickens', etc. -- is no longer protected by copyright.
You could not, however, scan the book and distribute the images of the pages. Because although the original author's text is not under copyright protection, the book itself (layout, design, etc.) could be. Also, any changes they might have made to the text (new grammar, or diction) could be, which is why you'd have to be careful. I think it would only qualify as a new protected work if the changes represented "an original work of authorship" according to 17 U.S.C. 101, but depending on the publisher they might try to sue you into bankruptcy anyway.
As long as you didn't copy the layout or any of the additional materials (critical essays, introductions) that publishers put into re-prints of classic literature, I see no reason why it would be illegal to type in and share the $2.99 Penguin Classics edition of Tom Sawyer that you can get at any Borders.
No, that's the point of a NAT router with a firewall.
A device doing pure NAT would forward all traffic -- including unsolicited incoming traffic -- on the public IP of the router to your computer's internal IP. So yes, you're right, someone on the outside wouldn't know your computer's internal IP address, but it wouldn't matter because the NAT box would happily forward all the packets. Because that's what NAT means.
Most home router boxes have NAT with selective port forwarding, plus a stateful-firewall-esque feature where it doesn't forward packets that aren't part of a connection that was initiated from inside the network. These are strictly speaking not part of NAT, they just tend to be rolled into the same devices making people think of them as one unit. (On most routers you can turn these things off and get "pure NAT" by putting your computer in as the DMZ address, this gets you NAT without firewalling so you can run a server.) But all of them could be easily implemented on IPv6 with a firewall.
The perceived security of NAT comes from the firewall-type features that are built into most home routers, not from the address translation service itself. There's no security gained simply by not letting an outside attacker know your internal network address if your gateway passes all packets, and conversely none given away by letting your attacker know it, IF you have a good firewall that's set to reject unsolicited traffic.
With IPv6, users would still probably want to have a box sitting between their easily-owned Windows box and the public net, but instead of being a NAT+Firewall, it would just be a Firewall. To them, the use would be exactly the same, except that they would have real end-to-end connectivity when they wanted it, and still retain the same level of security (or lack thereof) that they have today.
Now this is not a bad idea at all.
I think people are reaching too far, envisioning whole-house DC lines and stuff that wasn't a great idea when Thomas Edison was pushing for it, and it's not any better today. You don't need low-voltage wiring running all over your house, because most of those low-voltage devices are clustered together. Lots of them in the computer room, maybe a bunch in the entertainment/TV room, maybe a few more in the bedroom. Not too many in the garage, very few in the workshop/laundry/utility room. Plus there are lots of things which still need AC (HVAC blower, fridge, oven, clothes dryer, anything that makes heat, almost all lighting) and you'd be wasting a lot of copper running twice the lines. It's overkill to run wires throughout the whole house when a few strategically placed rectifers could do the job.
However it is equally ridiculous to have each device make its own DC, if they all want the same voltage (or close to it). Personally I think 24 or 48VDC would be a good standard, because then you have headroom enough to use a DC-DC converter to step down or regulate the power to whatever the device needs. For computer peripherals I can't think of anything that wouldn't be able to run off of that voltage except for CRT displays, and they're going away anyway.
I would be all for a standardized, DC power connector for computer and entertainment equipment and peripherals. What I'd like to see is some sort of connector where the physical size of the connector is proportional to the current draw from the equipment. This is something I once saw on an old theatre lighting system: the receptacles were actually two buss bars, spaced an inch or so apart; the plugs were rectangular blocks with metal on the two long opposing sides, so that when you plugged them into the gap between the buss bars you'd connect to them. But the neat feature was that the plugs were longer if they drew more current: a 20A circuit would be say 4" long, and a 10A one only 2". If the house mains was only capable of supplying 200A, then there would only be 40" of space to plug into. It makes it very hard to overload the mains. They are also a bit scary to work with, since it's possible to stick fingers/objects/etc into the gap between the buss bars and create a short circuit across the entire incoming mains (which might come unfused right from the power company's transformer), and I think this is why they're not used anymore. OT: If anyone knows what this type of connector was called, I'd be interested to know.
Anyway, if I was going to design a DC power plug standard it would be with some system like that (but less apt to kill you), maybe with the recifier output being through two very narrow parallel slots, with some sort of a flexible insulating barrier keeping objects out. But the key thing would be that a high-current device (like a LCD monitor or CPU) would use a bigger plug than a peripheral, and the PS would only have enough room for as much current as it could produce. (And yes, you'd have to make 'plug splitters' illegal under a system like this.)
Converting between high and low voltages with DC power is much more difficult, and requires more complex equipment. (An AC transformer is two pieces of wire wrapped around a chunk of iron.)
It's my understanding too that even a good DC/DC converter is significantly less efficient than a transformer, so that the advantage you'd get by transmitting high voltage DC instead of AC would be lost when it came time to step the voltage back down to something useable.
DC is something you want to make at the point where there are a bunch of things that all want the same voltage. If all the equipment in your datacenter wants 12 and 24VDC, then go ahead and have one big (or several redundant) power supply for the 'center and distribute the DC. But if there are a lot of things which want different voltages you're better off sticking with AC and doing the voltage conversion before you rectify it.
I could easily see having a big power supply for a whole rack of CPUs, if all of them are the same type of motherboard and take the same voltages (+/-5, 12, etc.) but then all you've done is essentially DIYed a really space-inefficient blade server.
The other question is an economy of scale one: I'm not sure that the demand for large DC power supplies will be large enough, in the near future, to make them cheaper per watt than small (less than 1000W) ATX ones, especially if you want redundancy, at least up until you start talking about very large supplies like the ones for IBM z/Series mainframes and the like.
It was on the afternoon 'news in review' (Afternoon Edition?) a few days ago as well, plus there's been meta-coverage of the rootkit in stories about the amount of negative press Sony's received as a result of the rootkit and about the CA and NY lawsuits. I wouldn't be surprised if the stories about the lawsuits would have gotten coverage in business oriented publications (WSJ and others) although I don't subscribe to them so I can't check for sure.
From what I understand of the DRM scheme, once you install those kernel extensions, it keeps you from ripping the CD from there on out.
But the obvious flaw is that most users are going to rip the CD before they get around to listening to or checking out the additional content -- I know quite a few people who have iPods and their Macs are set up so that when you put in an Audio CD, iTunes is automatically launched, and when iTunes detects a new music CD, it automatically rips it into their Library. So by the time they'd get around to looking at that Data volume with the DRM installer, the music is already on their iPod.
I think the end result of this setup is that people will accidentally install the DRM after the fact, and never notice until for some reason they want to rip it again (maybe years later, if they haven't done a clean install of their system) and not be able to -- and I think to most people their first response will be to assume that the CD is scratched or otherwise bad, not that the computer is (since it rips all other music fine, and even ripped the CD fine originally).
Yeah I was about to say the same thing, you beat me to it. I think it's more likely that the kids will have the admin password and thus keep their clueless parents from messing up the computer. Guess it depends on the age of the kids and parents involved. Maybe younger parents with more computer experience with young kids would actually be in charge, but there are still a lot of families out there where the parents involvement in computer maintenance is limited to handing over their Visa card once in a while.
Anyone want to get a petition going for a Linux version?
Or is someone going to have to reverse-engineer the binary and produce a OpenRootkit that can automatically compile and execute itself on all POSIX-compliant systems?
If the application is just an executable that stands alone, generally there is not a permission dialog. You can copy that to anyplace on your file system, no problem at all. The exception is if you are a non-administrator user and you try to copy it to the Applications folder (which by default non-admins do not have write privs to).
Where you see the 'enter password' dialog consistently is when installing software using an actual Installer package -- usually you will not be allowed to go past the first step in the "Wizard" type thing (which just describes what the package is) without authenticating. In general, the Installer packages are only necessary when an application requires files to be put in various places around the system, especially in the Library, and isn't just a standalone application bundle. I have definitely seen programs which could have just been distributed as an application on a disk image using an Installer, but they're the exception to the rule (in this case, laziness on the developer's part produces the desired result, since most people won't make an Installer package if they don't have to -- why some people do, I don't know).
I've seen this but only on the Selectric I, and only on certain models -- the keyboard has the same number of keys as the later versions I believe, but the caps are labeled differently and the top row goes "~ 2 3 4 5 6 7 8 9" and then some special keys which vary. The one I've seen that was set up this way had some special keys on the one to the right of the numeral nine, for typing Library of Congress index card special characters that I've never seen on anything else. (A special type-ball was required also.)
I've never been able to figure out if the eight-number top row was standard on all of the very first generation of Selectrics, or was something they only did on special models, in order to squeeze special characters onto the keyboard that otherwise would have been left off.
No, other way around. Even if the EULA says that you can't resell the software, at least in some jurisdictions in the U.S. you can go ahead and legally do it anyway, because of what's known as the doctrine of "first sale."
Microsoft would probably grumble about it, but there's not a lot they could do, since the courts have ruled.
Basically I think the point of the original post was "So, what?" Meaning that this isn't some huge groundbreaking deal -- all that's new is that some company got MS's permission to resell software licenses, with the implication that MS is getting a cut, and people are acting like it's some big blow to the Evil Machine, when in reality people elsewhere could sell used software already, with or without permission.
That at least was my understanding of the implications of the comment.
My understanding was that IPv6 plays nicely with DHCP-like systems, because of its hierarchical nature. So your ISP would give you an address where the last portion was the MAC address of your router/gateway, and it would assign addresses out to your computer/fridge/toaster/etc. that had those units' MAC addresses in them. So the numbers are assigned from where a NAT gateway used to be, but they're globally (in theory) routable.
;)
I'm not getting that information from the official spec or anything, it's more the informal discussions I've had with people who think it's a good idea, so feel free to take it with a grain of salt.
You can be sure that if I thought there was going to be any sort of creepy central control or registration involved, especially with the U.N. or something like it, that I'd be the last person on the planet to adopt IPv6, regardless of the technical advantages. However that's the first I've heard of that line of thinking so I'm not about to start oiling up the assault rifle and breaking out the Spam quite yet.
I don't think that people do this over gasoline (well, somewhere, somebody probably does) but they sure do it over other products. There's the never-ending, epic battle between Ford and Chevy enthusiasts, for one. I've met photographers who were as zealous about Nikon camera equipment (and eager to convert everyone else) as any Mac or Windows user. (Not to mention Kodak film vs. Fuji...) Videographers go back and forth on whether Panasonic cameras are a better deal than Sony, or if the latter are worth the price. In most gun clubs or stores you can get a spirited argument going by suggesting that Smith and Wesson handguns are superior to Colt's. At a cooking school you could probably get your ear talked off as to whether German or French chefs knives are better, and within those which brands are best. I could go on and on.
The quick answer to your question is 'yes.' Whenever you get people who spend a large percentage of their life in one industry, they develop preferences that seem obsessively odd to outsiders. It is our own fixation on computers that makes us think that people aren't just like this about other things; but being a "geek" isn't restricted to computers, we just don't use that term for people whose interest goes towards other things.
I'm not sure that they'd have to, because the addresses aren't given out randomly, they're hierarchical.
So really they'd only have to try a few million addresses that are sub-addresses of your ISP, at least this is based on my understanding of how the protocol is going to work.
But still you have a point, it would be like trying to find someone's Ethernet MAC address by exhausting that entire solution space (at least, since the IPv6 address will contain the MAC one).
It's not much of a cartel -- you will always have companies like Speakeasy, who are willing to toss in a few extras in order to siphon some customers off of Comcast or the local DSL telco and up to a higher price point.
As long as we don't let one company acquire and control all of the "last mile" bandwidth distribution networks, and we keep the cost of entrance into the ISP market reasonably low, you won't ever have that tight a cartel. Especially once you get the market fully penetrated and the flow of new customers starts to taper off, companies will start to think about how they can take them from each other, and that's when you'll see the features become a selling point.
I think the threat of a cartel is somewhat overstated, as long as we're vigilant about keeping the big telcos out of bed with the regulatory apparatus. (Or getting them out of bed, as the case may be.)
there's no need for NAT, and so there will be no NAT for IPv6
You obviously have never seen some geeks around here. There is no need for IP over carrier pigeon, and yet there is.
NAT is common enough an idea and -- despite the fact that I find it kludgey -- people seem to like it enough that I have no doubt that people will implement it on IPv6. It's not like it would be incredibly difficult or anything, especially if you dedicated a computer to doing it (as opposed to demanding it in some small-format dedicated appliance). Sure, maybe you'll never be able to go out to Best Buy and get an IPv6 NAT box for $30, because there won't be that level of demand for it, but for somebody willing to roll their own I'm sure it'll exist.
This assumes that there's no competition between ISPs. Eventually you'll get some value-added DSL or wireless internet provider that will give you that feature, and in time if there's personal demand for it -- which I think there will be, because as interactive content becomes more pervasive people will want the end-to-end communication ability that you can't get with NAT -- it will spread.
I think the reason why ISPs don't offer static IPs right now also has to do with an assumption that people who get static IPs will run their own servers, which tend to be high traffic and put a lot of load on the network. So that "static IP fee" is really being used to subsidize traffic-handling, it's not just paying for that address (the cost to the ISP of which is essentially zero).
Although I'm not saying that ISPs don't gouge the consumer pretty hard, I don't think that they have enough of a cartel to sink a feature like personal static IPs, if it was technically feasible at little to no cost.
But the more appropriate analogy is: You don't take
your car in for complete engine rebuild if the engine
is running fine.
While this may be true for your car, it's definitely not true of a helicopter, or a generator at a power plant, or any other important piece of machinery.
Would you still fly on an airline if that was their attitude towards maintenance? "Nah, we're not going to tear down that turbine...it hasn't failed yet!"
I think perhaps you should reevaluate the importance of the Internet to our society today. I think we've well surpassed the relative importance of a car to an average driver.
Neither of these points are really arguments for the current system, if anything they're good arguments against it, and in favor of IPv6.
#1 is nothing but a direct consecquence of the current shortage of IPv4 addresses. Under IPv6, there'd be no reason why every device on your network couldn't get a separate "real" address. The way they're handed out -- using a hierarchy instead of finite blocks -- would allow your ISP to let your home DHCP router hand out globally addressable IPs if it was set up correctly. Assuming your ISP doesn't suck, that is, and that's really not the fault of the IP system, one way or the other.
#2 is pretty frightening, because it shows a misunderstanding of what NAT is and a certain amount of laziness about security in general. That said, there's no reason why you couldn't get a 'firewall in a box' that would provide just as much (or as little) security without the NAT facility. It's just that right now when you go and buy a "home firewall" from Linksys, it almost always includes NAT by default (because of point #1, the pressure by ISPs on home users to only have one IP address due to limited supply). There's no reason why this needs to be true, however, and the security comes from the firewall effect and not the address translation itself.
I beg to differ. I question whether you're serious or a troll, but I'll respond anyway and give you the benefit of the doubt.
Lots of companies which are big enough to have their own Class-A allocations assign all of their clients globally routable addresses. I can tell you this from personal experience.
They don't use personal firewalls, obviously, and I have no idea why you think this is related. Using a personal firewall at the client level has nothing to do with IP address allocation or NAT. You can assign every user on a subnet a globally unique IP address, and then still use a stateful firewall for security. This is what these companies do: you get the benefit of not having your applications negotiate NAT with the protection of firewalls separating the internal networks at various facilities from the global network.
As far as the cost thing, if you're big enough to have a Class A block, you're not paying individually for IP addresses, so there's no difference in cost between a client that has a unique address and a NAT one. In fact the NAT one is probably slightly more expensive because the NAT routers are probably more maintainance and support-intensive than a straight firewall.
In short, I don't think you know what you're talking about. You might be correct when it comes to small or medium businesses, who are buying their connection from an ISP who is going to charge them more for a lot of static IPs than a few dynamic ones that they can use with NAT, but this issue isn't relevant to IBM, Ford, Apple, or the rest of the Class A companies.