Frankly, I don't know why anyone bothers with coming directly into the US by land or sea if the screening methods bothers them.
Do you know how hard it is to swim across the Atlantic Ocean and then hit the southern border?
But I basically agree (and I have gone through that screening twice up to date) and honestly I even find it a bit funny. I always wonder if the officer on the other side also thinks it is ridiculous, because why should a terrorist be afraid to be easily identified after he committed his act (he is probably dead by then). Don't tell me there is a big database of all potential terrorists which is in any way reliable. So it is more or less security theater played to calm down people.
The only way to assure net neutrality is to encrypt every packet and randomize the ports on all new network protocols. This is true right now for some P2P and skype.
Given the current European policy on data retention, we should do it even for mail and instant messaging. Of course you should use sftp instead of ftp and ssh instead of telnet, and your SMTP sessions should go encrypted, but that is not enough. We should rewrite every protocol and make it look like IPSEC.
If you can decrypt the packets why do you assume the ISP can't? Especially if the key is included in the software it is pointless, if it is in the data stream the ISP can also get it. Also how do you intend to connect to a well unknown port? Portscanning and trying to guess which encrypted packet might be HTTP?
Also your mail won't be safe at all. Yes, the ISP can't read it, but at least in Germany all the mail providers have to have a real time tap for law enforcement and I believe that is fairly standard all over the world now because of all those evil terrorists mailing their plans all the time (at least it seems politicians think so). So you need to encrypt your mail protecting the contents, but the receiver is still plainly visible as it would otherwise be impossible to foreward it to the receiver. And as far as I can see, encryption of mail basically failed or how often do you send encrypted mail?
Looking for technical solutions to social problems won't work. And I don't see it working for political problems either.
Session tokens appear in the referrer. So on a forum which allows user controlled images I can just craft an image which makes a request to my server, gets the referrer with the security token and then sends a HTTP redirect response (e.g. 302) to tell your browser to visit a different site. It is primitive to add the correct security token and your browser sends the right cookie. So the attack is more limited, but obviously the security token idea will only work partially.
It will also work on any site having ads when I can get into the ads stream. You need to rewrite the session token for each and every request. But I don't see an easy way to then allow multiple frames to be open with that site. Also caching of pages is no longer possible.
One other way is to ask for confirmation on state changing requests. But honestly do you want to see "Are you sure" all the time?
No, it prevents the production of an uncommented text. It is the original text and (I hope, I never read it) you should also be able to tell the text and the comments apart.
But yes, it is uses the copyright in a special way. If it is for political reasons is a lot harder to say. The original intention was to ban the book completely, but in the 50's some people demanded it to be reproduced again, so people could have access to it, but not reproduced in original, but in a commented way.
I think, the fear about remaining Nazis was still big enough. It isn't like they just vanished after 45, a lot of the minor ranks and obviously the biggest parts of the supporters were still around. Also quite some of the medium to higher ranks managed to camouflage themselves and so remained back. In the run of 60 years a lot of scandals surfaced, e.g. some of the medical doctors who experimented with the lives of prisoners in the KZ were later important figures in medical associations, a lot of the judges remained and a lot of the rest of the bureaucracy remained. It took literally decades to finally get rid of the brown scum, often scandal after scandal.
Also the often so called "censorship" is a bit more complicated. It is in Germany not allowed to use the insignia of the Nazi-regime. The exception is in a historical context and that goes quite far. E.g. the Indiana Jones movies were all allowed to remain unchanged as it was deemed historical context, even as it is fully fictional.
Sorry, but that is completely wrong. "Mein Kampf" is not banned, you can own it, you can read it. But the copyright of it is in the hands of the state of Bavaria and has been changed to last till eternity. The state of Bavaria refuses to publish an uncommented version of it. But you can get commented versions, e.g. on Amazon.
Make ALL personal information your personal property, the use of which is revocable at will, like the RIAA does with copying music. Anyone you aren't doing business with (say, Choicepoint, Lexis/Nexis, USSEARCH.COM etc.), who is trying to share your personal information around, has to ask for permission and pay royalties for transactions.
Apart from the royalties, that is what the EU privacy regulation provides. I have the right to know from any holder of private information about me, what he holds, if not willingly by the company then by the help of a judge. I have the right to have my data deleted when the holder has no longer any business relation to me. One of the basic requirements of those laws is, that a company may only collect information it needs and only stores information it needs. Without an express permission by the customer a company is not permitted to pass on personal data.
E.g. the IP to customer ID for flatrates have to be deleted. At least that is the idea. And here in Germany the biggest ISP (t-online) actually lost in that regards (they kept it for 180 days) and had to install software to delete the data for the one person who had sued them. As more and more people threatened to sue they now changed their policy to store that data for "only" 7 days. It might have been in vain (there is a new law directly opposing it, for "security" concerns that data soon has to be stored for 12 months).
"What's that, T-Mobile won't let you talk to VOIP users? Come to OUR phone service. We don't cripple our phones. You can talk to anybody."
"And you only have to pay ten times the rates". If nobody uses the service they make money of, they will increase the rates on the data service. And what have you gained then?
On the Wi-Fi network. Here in Germany nationwide frequencies were auctioned by the state and some providers are building networks. The traditional cell companies were not even interested. All of them had played with prototypes. After the debacle with 3G they seem to have become a lot more careful. They obviously calculated what a new infrastructure would cost them. And they already have quite some of the infrastructure in place and would "only" need to install new antennas and some electronics. Somehow I have the impression they found something with the prototypes which didn't make Wi-Fi interesting for them.
Try to create a new account with that e-mail address. See if you get an error message saying the address is already associated with an account.
Log in under an existing account, and try to switch to another e-mail address. See if you get an error message saying the address is already associated with an account.
Use the forgot-your-password feature to request a password be sent to a given e-mail address. See if you get an error message saying that address is not associated with an account.
The first two points of the list are based on a very much flawed assumption by the applications, the problem to get a mail-address. The third is just plain stupid and leaks information.
Why are the first two flawed? They assume, that it is hard to get a second mail-address and therefore allowing only one account to be associated with an email-address somehow makes it harder for people to sign up multiple times. If that is not the reason, then the only other one I find is, to collect as many email-addresses as possible.
For me, there is no real reason, why an email-address can't be used multiple times. If you are afraid that someone signs up thousands of accounts, limit the number of signups per week. The probability that someone else trying to find out, if the address is being used hits one of the weeks where you signed up an account is much smaller.
Third and probably the most important is that the EU and the rest of the world have only been attempting to reduce Co2 emisiosn since 2000 when the kyoto accord was in effect. Comparing to anything previous is senseless and misleading. It implies there was an effort that isn't and attempt to say look, we are guilty because we done this before that.
Sorry, but you are wrong. First energy preservation and reduction of emissions started here in Germany in the late 80's, not because of the climate change, but because our woods were dying. That it also reduced emissions is a side effect, true, but it nonetheless did reduce them. Second, the EU has made a claim to go beyond Kyoto, the targets the EU has set itself is 20% lower than the emission levels of 1990. Kyoto was not the start, as it was known before that the CO2 levels were rising and preparations of that conference also took some time. Also the decline of the economy especially in Eastern Germany with the big brown coal power plants had a big impact on the CO2 levels. Sure, not a direct effect of the effort to reduce the emission, but it sure helped.
Also why did you take 2000 as the date of the Kyoto-protocol? According to Wikipedia it was opened for signature in 1997. And the Kyoto protocol always compares the emissions to the standard of 1990.
But I also have to say, I don't like, that the emissions rose in the last 7 years. Even 1% more means that they rose and the target is clearly to reduce them. Seeing some of the nonsense the German administration has done in the last time (e.g. opposed a clear regulation for the reduction of emissions of the fleet produced by car makers (thanks to heavy lobbying)) is not a good sign. I don't think in a global effort it is good to always excuse your own faults by faults others made/make. It doesn't help.
In every press release AMD stated it will run in AM2 sockets. If I remember correctly it will not be able to use the new hypertransport links, support for the new power saving functions (it can switch off complete cores if they aren't needed) in AM2 sockets, it will need AM2+ for that. Sorry, I am far too lazy to search for a reference for those last bits of information, it is something I read in a magazine (paper version).
Where have you learnt C++? If you use iostream you can set them to throw exceptions: Exception Mask. It isn't done by default, because a failing IO-operation is hardly a big surprise (incorrect rights, nonexisiting file, out of data). I find exceptions overrated, they make it awkward to check the state and should be reserved where you really need to unwind the stack not for simple state checks.
To get free advertising on the internet. And even when 99% would turn out to be crap to at least get some of decent quality to air them on TV. Afterall who hasn't heard of some videos on YouTube and other of those sites which have fantastic quality. So obviously you can get people to do the videos and by doing so get them to also do the advertising.
It just shows again how much people in big corporations without any clue (or do you believe the manager who had that idea ever had visited YouTube?) trust the bullshit of some consultants and parties with vested interest in promoting that stuff.
As far as I understand it, the assumption was, that when the lead in two probes was identical then the parts belonged to the same bullet. They just bought some of the same ammunition (I have to admit I am rather surprised you can still find such old ammunition) and found out, that two or more bullets can share the same material (which honestly isn't very surprising if you think about it, I would assume most of the bullets made from the same lead batch are basically identical for chemical analysis and also will probably end up in the same box).
At least that is how I read the article. What they say is, that there could have been more bullets involved than the original report gave credit for, because it was based on a flawed assumption. They can not prove or disprove that, as they didn't have the original evidence and only used the same kind to find out, if the basic assumption was right or not.
Actually that data is only accessible by the office (called Birthler Behörde in German after its head, because the official name "Bundesbeauftragte für die Unterlagen des Staatssicherheitsdienstes der ehemaligen Deutschen Demokratischen Republik" is a bit long) who only grants access to the people, who were spied upon to find out, what was known and who did it. Journalists and researchers can request access and there is a process in place where the victim can deny the request.
There is an additional process to query, if someone was working for the Stasi, but without revealing details, for some jobs that gets queried.
You know what would be totally ironic? When the ads became such a big nuisance that even the biggest couch potato would be annoyed and would stop watching and do something else instead. Perhaps we should wish for the networks and ad companies to over do it?
Nothing in general (although I am not so sure I like that business), but the owners of the rights currently sell to the networks, which on their part pay them and then recoup the money with ads (or subscriptions). So their business model up to now was to sell the rights to someone for a fixed prize and I don't see them willing to get into the business to compete with their primary customers. That would seriously limit their ability to negotiate (the big media companies will claim the value is smaller, because they no longer have exclusive control and so on).
I also don't claim it wouldn't work, I just have the feeling that those companies wouldn't want to do it, they are not used to it and I think they would want quite some control on which ads to appear.
There is a big problem for them, though. They don't have any interest to provide their 'content' for free, they want to earn money. But honestly would you pay for something of the quality of YouTube? I fear most people would want to get a better quality, if they have to pay for it. I seriously doubt that they are willing to go for advertising supported distribution, because of the hard to calculate revenue.
And that gets them into this problem, the net was for very long not able to deliever high quality video streams (and perhaps it still isn't when a lot of people want to watch it). And I don't see any way for them to compete with YouTube, if they deliever that low quality people will (rightfully) ask why they should pay, if they deliever high quality they get into serious problems of where to get the bandwidth and how to pay for it.
So I can see, that for them the only way is to make sure, that their content has a high enough value to either make it possible to sell it to YouTube or to at least make a lot of money by selling to the traditional channels.
For your information, here in Germany the DFL (German Football Leage) who has the rights and sells them last time also sold separate rights for delievery via IP and mobile phone and you can bet, that the owner of the rights will sue anyone 'broadcasting' it for free.
Now could someone please explain to me what cross site scripting is and why it is so hard to stamp it out.
One of the limits on active content (mostly Javascript, but it also applies to some others) is, that a script can only access pages originating from the same server. So a script from server A can change a picture or hide parts of a page which is also from server A, but it is not allowed to do that to stuff comming from server B.
A cross site scripting vulnerability now enables an attacker to do exactly that, add some script not from server A which is presented in a way, that your browser believes it is from server A and allows it therefore manipulations which are unsafe. That basically means, someone is able to inject arbitrary code into pages and do operations in their context.
There is a way to not have cross site scription (shorthand for that is XSS) vulnerabilities, white listing. White listing means, you only accept known good data in anything which reaches the server. That includes uploads and forms, URLs and cookies. The problem is, that most sites don't do white listing, but the opposite, black listing. Black listing means, you try to find the things which will break your site and remove them. The problem is, that you will never know all ways which will break your site.
So why isn't white listing used exclusively? It is much harder, you have to specify what is acceptable, so you are less flexible, whenever you add something as acceptable you have to be sure it won't break something. Black listing is much easier to implement and things you didn't think of work, with the downside that also bad things you didn't think about sometimes work.
If I've got an RS-232 cable plugged into a computer and its nothing more than few lengths of copper wire, then I run a piece of software on the computer and now there's electrical current on pins 2, 8, and 11 of the cable, I'd say that's caused a change in the physical world. It's not just a "mental process," any more than a [patentable] method of putting threads on a screw is a mental process.
And I delete your software, write one of my own and also make pins 2, 8 and 11 change their electrical potential. And now tell me, is the ability to exhibit reactions to the physical world a property of your software, of my software or perhaps of the common hardware running the software?
One of the really nice additions is the geometry shader. As far as I understand it, MS won't port it to DX9 and it will probably have a big impact. Up to now you only had the vertex shader which couldn't create geometry, it could only move vertices around or change the normals of a face.
The geometry shaders can actually create geometry and that means, that a complete new class of algorithms are possible to run on the GPU. E.g. a vertex shader can animate a character with vertex morphing or a skeleton system. But it can't add vertices on demand, so you either have a very high vertex count or have to live with deformations. Also it can't do level of detail and also can't smooth out a character as all of those mean you have to add new geometry.
Another example would be landscape or plants. The geometry shader in theory should be able to create the complete landscape from a height texture. The basic algorithm for that is simple (each pixel in the texture represents a height, the x and y coordinates are just a rectangular grid), the problems there is again level of detail. The last batch of algorithms tend to fire off a large amount of vertices to the GPUs memory and then compile index arrays of which vertices to use on the CPU and fire that to the GPU. The geometry shader should be able to do that all by itself, further unloading the CPU. For plants the situation is similar, many plant system are procedural in nature, you can literally grow the plants. Currently they are often precalculated and then stored in graphics card memory. With a fast geometry shader you could actually run the generating algorithm and so create thousands of different trees on demand.
The first batch of DX10 capable cards won't be able to do all of these, as their power will be limited, but they promise to offer a new way to evolve the capabilities. The vertex shaders were quite limited and often they were not even used. Or they were used to precalculate data for the fragment shaders (sorry, but I use the OpenGL terminology, I think in DX terms it is pixel shaders), so they worked on dummy vertices and created a texture map which was later used in the fragment shader to actually display something.
I don't see, why it shouldn't be possible to add the geometry shaders to DX9, as they are basically only a new kind of shaders you have to pass to the GPU. But I can't imagine MS to port them, at least not if the industry doesn't swing to OpenGL to use the new features and Microsoft has to fear that DX might get damaged.
So I would expect the new cards to be able to handle considerable more dynamically created geometry. The other part is, that as was already posted in another thread, DX9 is quite inefficient to post small objects to the GPU. A very nice way to speed up rendering is to create an array of all the data (DX uses the term "buffers" for them, in GL lingo it is "Vertex Arrays") and then tell the driver to use the data in those to render. The nice thing is, you can also allocate space directly on the graphics card and then to render one of those objects you don't need to pass the geometry from main memory to the graphics cards memory. DX9 has a problem when those arrays are small and don't contain a big amount of vertices, it is then actually slower than sending the vertices one by one. DX10 is said to change that and make very small arrays useful.
Have you checked the images on a screen? My cameras display also has big problems displaying violet, but it captures it ok. On a normal computer screen it is violet, but the tiny display on the camera seems to have a big problem with blue and especially violet and so it looks much too blue. I therefore don't use it anymore to decide if the color is right and just trust my experience.
Germany, for example, gets by with a per capita energy consumption of around 40% of that of the United States without a significant loss in standard of living. How is this done? People habitually turn off lights in rooms that they are not in; smaller, more fuel efficient cars are the norm;
The density of SUVs is still lower, yes, but I wouldn't call the cars very fuel efficient either. Additionally the lack of a speed limit means that the average speed is quite a bit higher resulting in higher fuel consumption.
waste products are heavily mined for reusable resources; every major city has a reliable, efficient, and widely utilized public transportation networks;
the "heavily mined for reusable resources" has a downside as well. As most plastic is removed from general waste (as are metals and glass) the rest doesn't burn well. So waste incinerating plants nowadays actually have to add oil to even work, not exactly a sensible development, because the energy needed to rework the recollected plastic isn't low.
The public transportation system is far from perfect. It basically works in most cities, but the railway to connect them isn't good. There are some very fast connections (I live near a new track where 300 km/h in regular
traffic is reached), but many are quite desolate and need work. A lot of traffic is still road traffic.
people tend to choose bicycles or walking to nearby locations rather than driving; individuals reuse packaging (you bring your own bags to the grocery store); products are generally packed in less packaging material. Some of these things are done by individuals, some of them require government or corporate intervention. However, millions of people choosing to do the right thing creates significant, measurable results on that country's energy footprint.
At least one big measure is missing, houses are insulated. There are mandatory insulation values a new house has to reach. If it is loosing too much energy you have to improve it. Windows with two glass panes are the norm since the late seventies, many buildings nowadays have three layered glass windows reducing energy loss through the windows. Windows and roofs are the parts where many older building loose most energy.
There are also some buildings needing no heating or cooling at all. The are called Passive-House.
Honestly all they need to do is make the template engine scrub any script that does redirects or nasty tricks like opening popups on load.
If you find a way to do that, you will also have solved the halting-problem, in other words, that is nearly impossible to do.
There is only one way which might be safe, supply finished javascript functions to the users to use and make it impossible to define new functions. Even that might be dangerous.
I am not so sure about that, though. The author seems to be from Great Brittain, so he probably could sue before a British Court. The website is accessible in GB and so the MPAA is violating the British copyright.
But honestly I think we shouldn't step down to the low standard of the MPAA. I think firing off the email and making it public is probably the right way. Afterall the MPAA relies on its public image and if more and more details surface about them not being a honest party it might inflict much bigger harm to them in the long run. In contrast to all the public traded companies the MPAA represents we don't have to fullfill a 3 month plan to make money and so we can wait.
I think we should give the MPAA enough rope to hang itself. The RIAA already looks rather tangled up in wires they spanned themselves.
AMD's actually guilty of the same flawed logic though - their criticism of Intel's 4 core processor being just 2 dual cores stuck together is just as pointless. It doesn't matter what matters is how well the processor meets the requirements of its target market.
But AMD is right, it is no quad core, it is a multichip package of two dual cores. So calling it "quad core" is pure marketing speech because INTEL is lagging behind AMD again. They simply don't have a real quad core yet. So they use those multichip processors to hide the fact that they behind and even try to create the illusion that it is AMD which is behind.
Realistically you have to compare an INTEL quad core with an AMD dual processor dual core setup. But that is not fair either, because now the AMD rig has two independent memory systems and so has a big advantage on big data sets. But it also needs much more space. So basically the multichip module from INTEL only saves you space on the mainboard, technically it is a dual processor setup with a bad memory path.
And that could hurt AMDs real quad cores, because if everybody in the marketplace associates quad cores with INTELs multichip module they might severely underestimate the performance AMDs processor (hopefully) has. And that would hurt AMDs sales.
Slashdot Mirror
Do you know how hard it is to swim across the Atlantic Ocean and then hit the southern border?
But I basically agree (and I have gone through that screening twice up to date) and honestly I even find it a bit funny. I always wonder if the officer on the other side also thinks it is ridiculous, because why should a terrorist be afraid to be easily identified after he committed his act (he is probably dead by then). Don't tell me there is a big database of all potential terrorists which is in any way reliable. So it is more or less security theater played to calm down people.
If you can decrypt the packets why do you assume the ISP can't? Especially if the key is included in the software it is pointless, if it is in the data stream the ISP can also get it. Also how do you intend to connect to a well unknown port? Portscanning and trying to guess which encrypted packet might be HTTP?
Also your mail won't be safe at all. Yes, the ISP can't read it, but at least in Germany all the mail providers have to have a real time tap for law enforcement and I believe that is fairly standard all over the world now because of all those evil terrorists mailing their plans all the time (at least it seems politicians think so). So you need to encrypt your mail protecting the contents, but the receiver is still plainly visible as it would otherwise be impossible to foreward it to the receiver. And as far as I can see, encryption of mail basically failed or how often do you send encrypted mail?
Looking for technical solutions to social problems won't work. And I don't see it working for political problems either.
Session tokens appear in the referrer. So on a forum which allows user controlled images I can just craft an image which makes a request to my server, gets the referrer with the security token and then sends a HTTP redirect response (e.g. 302) to tell your browser to visit a different site. It is primitive to add the correct security token and your browser sends the right cookie. So the attack is more limited, but obviously the security token idea will only work partially.
It will also work on any site having ads when I can get into the ads stream. You need to rewrite the session token for each and every request. But I don't see an easy way to then allow multiple frames to be open with that site. Also caching of pages is no longer possible.
One other way is to ask for confirmation on state changing requests. But honestly do you want to see "Are you sure" all the time?
No, it prevents the production of an uncommented text. It is the original text and (I hope, I never read it) you should also be able to tell the text and the comments apart.
But yes, it is uses the copyright in a special way. If it is for political reasons is a lot harder to say. The original intention was to ban the book completely, but in the 50's some people demanded it to be reproduced again, so people could have access to it, but not reproduced in original, but in a commented way.
I think, the fear about remaining Nazis was still big enough. It isn't like they just vanished after 45, a lot of the minor ranks and obviously the biggest parts of the supporters were still around. Also quite some of the medium to higher ranks managed to camouflage themselves and so remained back. In the run of 60 years a lot of scandals surfaced, e.g. some of the medical doctors who experimented with the lives of prisoners in the KZ were later important figures in medical associations, a lot of the judges remained and a lot of the rest of the bureaucracy remained. It took literally decades to finally get rid of the brown scum, often scandal after scandal.
Also the often so called "censorship" is a bit more complicated. It is in Germany not allowed to use the insignia of the Nazi-regime. The exception is in a historical context and that goes quite far. E.g. the Indiana Jones movies were all allowed to remain unchanged as it was deemed historical context, even as it is fully fictional.
Sorry, but that is completely wrong. "Mein Kampf" is not banned, you can own it, you can read it. But the copyright of it is in the hands of the state of Bavaria and has been changed to last till eternity. The state of Bavaria refuses to publish an uncommented version of it. But you can get commented versions, e.g. on Amazon.
Apart from the royalties, that is what the EU privacy regulation provides. I have the right to know from any holder of private information about me, what he holds, if not willingly by the company then by the help of a judge. I have the right to have my data deleted when the holder has no longer any business relation to me. One of the basic requirements of those laws is, that a company may only collect information it needs and only stores information it needs. Without an express permission by the customer a company is not permitted to pass on personal data.
E.g. the IP to customer ID for flatrates have to be deleted. At least that is the idea. And here in Germany the biggest ISP (t-online) actually lost in that regards (they kept it for 180 days) and had to install software to delete the data for the one person who had sued them. As more and more people threatened to sue they now changed their policy to store that data for "only" 7 days. It might have been in vain (there is a new law directly opposing it, for "security" concerns that data soon has to be stored for 12 months).
"And you only have to pay ten times the rates". If nobody uses the service they make money of, they will increase the rates on the data service. And what have you gained then?
On the Wi-Fi network. Here in Germany nationwide frequencies were auctioned by the state and some providers are building networks. The traditional cell companies were not even interested. All of them had played with prototypes. After the debacle with 3G they seem to have become a lot more careful. They obviously calculated what a new infrastructure would cost them. And they already have quite some of the infrastructure in place and would "only" need to install new antennas and some electronics. Somehow I have the impression they found something with the prototypes which didn't make Wi-Fi interesting for them.
The first two points of the list are based on a very much flawed assumption by the applications, the problem to get a mail-address. The third is just plain stupid and leaks information.
Why are the first two flawed? They assume, that it is hard to get a second mail-address and therefore allowing only one account to be associated with an email-address somehow makes it harder for people to sign up multiple times. If that is not the reason, then the only other one I find is, to collect as many email-addresses as possible.
For me, there is no real reason, why an email-address can't be used multiple times. If you are afraid that someone signs up thousands of accounts, limit the number of signups per week. The probability that someone else trying to find out, if the address is being used hits one of the weeks where you signed up an account is much smaller.
Sorry, but you are wrong. First energy preservation and reduction of emissions started here in Germany in the late 80's, not because of the climate change, but because our woods were dying. That it also reduced emissions is a side effect, true, but it nonetheless did reduce them. Second, the EU has made a claim to go beyond Kyoto, the targets the EU has set itself is 20% lower than the emission levels of 1990. Kyoto was not the start, as it was known before that the CO2 levels were rising and preparations of that conference also took some time. Also the decline of the economy especially in Eastern Germany with the big brown coal power plants had a big impact on the CO2 levels. Sure, not a direct effect of the effort to reduce the emission, but it sure helped.
Also why did you take 2000 as the date of the Kyoto-protocol? According to Wikipedia it was opened for signature in 1997. And the Kyoto protocol always compares the emissions to the standard of 1990.
But I also have to say, I don't like, that the emissions rose in the last 7 years. Even 1% more means that they rose and the target is clearly to reduce them. Seeing some of the nonsense the German administration has done in the last time (e.g. opposed a clear regulation for the reduction of emissions of the fleet produced by car makers (thanks to heavy lobbying)) is not a good sign. I don't think in a global effort it is good to always excuse your own faults by faults others made/make. It doesn't help.
In every press release AMD stated it will run in AM2 sockets. If I remember correctly it will not be able to use the new hypertransport links, support for the new power saving functions (it can switch off complete cores if they aren't needed) in AM2 sockets, it will need AM2+ for that. Sorry, I am far too lazy to search for a reference for those last bits of information, it is something I read in a magazine (paper version).
Where have you learnt C++? If you use iostream you can set them to throw exceptions: Exception Mask. It isn't done by default, because a failing IO-operation is hardly a big surprise (incorrect rights, nonexisiting file, out of data). I find exceptions overrated, they make it awkward to check the state and should be reserved where you really need to unwind the stack not for simple state checks.
To get free advertising on the internet. And even when 99% would turn out to be crap to at least get some of decent quality to air them on TV. Afterall who hasn't heard of some videos on YouTube and other of those sites which have fantastic quality. So obviously you can get people to do the videos and by doing so get them to also do the advertising.
It just shows again how much people in big corporations without any clue (or do you believe the manager who had that idea ever had visited YouTube?) trust the bullshit of some consultants and parties with vested interest in promoting that stuff.
As far as I understand it, the assumption was, that when the lead in two probes was identical then the parts belonged to the same bullet. They just bought some of the same ammunition (I have to admit I am rather surprised you can still find such old ammunition) and found out, that two or more bullets can share the same material (which honestly isn't very surprising if you think about it, I would assume most of the bullets made from the same lead batch are basically identical for chemical analysis and also will probably end up in the same box).
At least that is how I read the article. What they say is, that there could have been more bullets involved than the original report gave credit for, because it was based on a flawed assumption. They can not prove or disprove that, as they didn't have the original evidence and only used the same kind to find out, if the basic assumption was right or not.
Actually that data is only accessible by the office (called Birthler Behörde in German after its head, because the official name "Bundesbeauftragte für die Unterlagen des Staatssicherheitsdienstes der ehemaligen Deutschen Demokratischen Republik" is a bit long) who only grants access to the people, who were spied upon to find out, what was known and who did it. Journalists and researchers can request access and there is a process in place where the victim can deny the request.
There is an additional process to query, if someone was working for the Stasi, but without revealing details, for some jobs that gets queried.
You know what would be totally ironic? When the ads became such a big nuisance that even the biggest couch potato would be annoyed and would stop watching and do something else instead. Perhaps we should wish for the networks and ad companies to over do it?
Nothing in general (although I am not so sure I like that business), but the owners of the rights currently sell to the networks, which on their part pay them and then recoup the money with ads (or subscriptions). So their business model up to now was to sell the rights to someone for a fixed prize and I don't see them willing to get into the business to compete with their primary customers. That would seriously limit their ability to negotiate (the big media companies will claim the value is smaller, because they no longer have exclusive control and so on).
I also don't claim it wouldn't work, I just have the feeling that those companies wouldn't want to do it, they are not used to it and I think they would want quite some control on which ads to appear.
There is a big problem for them, though. They don't have any interest to provide their 'content' for free, they want to earn money. But honestly would you pay for something of the quality of YouTube? I fear most people would want to get a better quality, if they have to pay for it. I seriously doubt that they are willing to go for advertising supported distribution, because of the hard to calculate revenue.
And that gets them into this problem, the net was for very long not able to deliever high quality video streams (and perhaps it still isn't when a lot of people want to watch it). And I don't see any way for them to compete with YouTube, if they deliever that low quality people will (rightfully) ask why they should pay, if they deliever high quality they get into serious problems of where to get the bandwidth and how to pay for it.
So I can see, that for them the only way is to make sure, that their content has a high enough value to either make it possible to sell it to YouTube or to at least make a lot of money by selling to the traditional channels.
For your information, here in Germany the DFL (German Football Leage) who has the rights and sells them last time also sold separate rights for delievery via IP and mobile phone and you can bet, that the owner of the rights will sue anyone 'broadcasting' it for free.
One of the limits on active content (mostly Javascript, but it also applies to some others) is, that a script can only access pages originating from the same server. So a script from server A can change a picture or hide parts of a page which is also from server A, but it is not allowed to do that to stuff comming from server B.
A cross site scripting vulnerability now enables an attacker to do exactly that, add some script not from server A which is presented in a way, that your browser believes it is from server A and allows it therefore manipulations which are unsafe. That basically means, someone is able to inject arbitrary code into pages and do operations in their context.
There is a way to not have cross site scription (shorthand for that is XSS) vulnerabilities, white listing. White listing means, you only accept known good data in anything which reaches the server. That includes uploads and forms, URLs and cookies. The problem is, that most sites don't do white listing, but the opposite, black listing. Black listing means, you try to find the things which will break your site and remove them. The problem is, that you will never know all ways which will break your site.
So why isn't white listing used exclusively? It is much harder, you have to specify what is acceptable, so you are less flexible, whenever you add something as acceptable you have to be sure it won't break something. Black listing is much easier to implement and things you didn't think of work, with the downside that also bad things you didn't think about sometimes work.
And I delete your software, write one of my own and also make pins 2, 8 and 11 change their electrical potential. And now tell me, is the ability to exhibit reactions to the physical world a property of your software, of my software or perhaps of the common hardware running the software?
One of the really nice additions is the geometry shader. As far as I understand it, MS won't port it to DX9 and it will probably have a big impact. Up to now you only had the vertex shader which couldn't create geometry, it could only move vertices around or change the normals of a face.
The geometry shaders can actually create geometry and that means, that a complete new class of algorithms are possible to run on the GPU. E.g. a vertex shader can animate a character with vertex morphing or a skeleton system. But it can't add vertices on demand, so you either have a very high vertex count or have to live with deformations. Also it can't do level of detail and also can't smooth out a character as all of those mean you have to add new geometry.
Another example would be landscape or plants. The geometry shader in theory should be able to create the complete landscape from a height texture. The basic algorithm for that is simple (each pixel in the texture represents a height, the x and y coordinates are just a rectangular grid), the problems there is again level of detail. The last batch of algorithms tend to fire off a large amount of vertices to the GPUs memory and then compile index arrays of which vertices to use on the CPU and fire that to the GPU. The geometry shader should be able to do that all by itself, further unloading the CPU. For plants the situation is similar, many plant system are procedural in nature, you can literally grow the plants. Currently they are often precalculated and then stored in graphics card memory. With a fast geometry shader you could actually run the generating algorithm and so create thousands of different trees on demand.
The first batch of DX10 capable cards won't be able to do all of these, as their power will be limited, but they promise to offer a new way to evolve the capabilities. The vertex shaders were quite limited and often they were not even used. Or they were used to precalculate data for the fragment shaders (sorry, but I use the OpenGL terminology, I think in DX terms it is pixel shaders), so they worked on dummy vertices and created a texture map which was later used in the fragment shader to actually display something.
I don't see, why it shouldn't be possible to add the geometry shaders to DX9, as they are basically only a new kind of shaders you have to pass to the GPU. But I can't imagine MS to port them, at least not if the industry doesn't swing to OpenGL to use the new features and Microsoft has to fear that DX might get damaged.
So I would expect the new cards to be able to handle considerable more dynamically created geometry. The other part is, that as was already posted in another thread, DX9 is quite inefficient to post small objects to the GPU. A very nice way to speed up rendering is to create an array of all the data (DX uses the term "buffers" for them, in GL lingo it is "Vertex Arrays") and then tell the driver to use the data in those to render. The nice thing is, you can also allocate space directly on the graphics card and then to render one of those objects you don't need to pass the geometry from main memory to the graphics cards memory. DX9 has a problem when those arrays are small and don't contain a big amount of vertices, it is then actually slower than sending the vertices one by one. DX10 is said to change that and make very small arrays useful.
Have you checked the images on a screen? My cameras display also has big problems displaying violet, but it captures it ok. On a normal computer screen it is violet, but the tiny display on the camera seems to have a big problem with blue and especially violet and so it looks much too blue. I therefore don't use it anymore to decide if the color is right and just trust my experience.
The density of SUVs is still lower, yes, but I wouldn't call the cars very fuel efficient either. Additionally the lack of a speed limit means that the average speed is quite a bit higher resulting in higher fuel consumption.
the "heavily mined for reusable resources" has a downside as well. As most plastic is removed from general waste (as are metals and glass) the rest doesn't burn well. So waste incinerating plants nowadays actually have to add oil to even work, not exactly a sensible development, because the energy needed to rework the recollected plastic isn't low.
The public transportation system is far from perfect. It basically works in most cities, but the railway to connect them isn't good. There are some very fast connections (I live near a new track where 300 km/h in regular traffic is reached), but many are quite desolate and need work. A lot of traffic is still road traffic.
At least one big measure is missing, houses are insulated. There are mandatory insulation values a new house has to reach. If it is loosing too much energy you have to improve it. Windows with two glass panes are the norm since the late seventies, many buildings nowadays have three layered glass windows reducing energy loss through the windows. Windows and roofs are the parts where many older building loose most energy.
There are also some buildings needing no heating or cooling at all. The are called Passive-House.
If you find a way to do that, you will also have solved the halting-problem, in other words, that is nearly impossible to do.
There is only one way which might be safe, supply finished javascript functions to the users to use and make it impossible to define new functions. Even that might be dangerous.
I am not so sure about that, though. The author seems to be from Great Brittain, so he probably could sue before a British Court. The website is accessible in GB and so the MPAA is violating the British copyright.
But honestly I think we shouldn't step down to the low standard of the MPAA. I think firing off the email and making it public is probably the right way. Afterall the MPAA relies on its public image and if more and more details surface about them not being a honest party it might inflict much bigger harm to them in the long run. In contrast to all the public traded companies the MPAA represents we don't have to fullfill a 3 month plan to make money and so we can wait.
I think we should give the MPAA enough rope to hang itself. The RIAA already looks rather tangled up in wires they spanned themselves.
But AMD is right, it is no quad core, it is a multichip package of two dual cores. So calling it "quad core" is pure marketing speech because INTEL is lagging behind AMD again. They simply don't have a real quad core yet. So they use those multichip processors to hide the fact that they behind and even try to create the illusion that it is AMD which is behind.
Realistically you have to compare an INTEL quad core with an AMD dual processor dual core setup. But that is not fair either, because now the AMD rig has two independent memory systems and so has a big advantage on big data sets. But it also needs much more space. So basically the multichip module from INTEL only saves you space on the mainboard, technically it is a dual processor setup with a bad memory path.
And that could hurt AMDs real quad cores, because if everybody in the marketplace associates quad cores with INTELs multichip module they might severely underestimate the performance AMDs processor (hopefully) has. And that would hurt AMDs sales.