Frankly, you're wrong, and this has nothing to do with tin foil hats. Microsoft is intent on making sure that Windows becomes the OS of your home entertainment system. If that means catering to the whims of the movie industry, I doubt Microsoft has any qualms about it. After all, if the movie industry doesn't like the protection Microsoft has set up for them, they'll just back someone else that will, and this would effectively push Microsoft out of the home entertainment system market.
heh. So true. Although I find it funny this article is named "FutureMark Confirms nVidia's Benchmark Cheating", when, in fact, FutureMark has determined that both NVIDIA and ATI are cheating at 3DMark03. If you read the section "What Is the Performance Difference Due to These Cheats?", they reveal the fact that ATI has cheated in their drivers optimizing the results of test 4 by 8.2%. Maybe this article ought to be titled "FutureMark Confirms Benchmark Cheating Prevalent Throughout Industry".
However, I do have some problems with the article. In several of them, it is mentioned that the vertex shader was replaced with a similar looking, but non-identical. Why was NVIDIA able to replace it with a similar but non-identical copy and boost performance so much? Is the 3DMark03 one too biased toward ATI cards, or are the ATI cards too biased towards 3DMark03? I believe these parts alone prove that you can try to make a benchmark try to reveal "real world" performance, but so long as there's multiple ways of doing the same thing without adversely affecting quality or appearance, it's difficult to believe any benchmark, game based or otherwise, is indicitive of real world performance.
There's also the issue of the fact that when those copyrights actually expire, will they be removable from the encryption? When the copyright term expires, they're supposed to be in the public domain, but if you're not legally allowed to break the encryption holding the work, how does it enter the public domain? We've created perpetual copyright without ever really intending to.
According to the research paper, Windows 2003 is up to 66% faster at providing Windows fileshares to Windows XP clients than Linux is. WOW! What a revalation! I would've never expected Windows to be able to transmit files with it's native protocol faster than the reverse engineered implementation, SAMBA.
Now, if Microsoft wants to provide full, complete and accurate documentation on SMB/CIFS and the MS-RPC function calls, maybe these numbers could be coaxed closer together... But I'm not holding my breath.
The points in that page are quite valid, however, not all configurations are necessarily vulnerable to an ARP poisoning attack. For example, many cable modems are capable of filtering which traffic is meant for them and which is not. In order to do ARP spoofing, you need to hear an ARP request to respond to, so if you never receive a request, you can't return a response. However, this isn't the case for all cable networks. In Winnipeg, we had both. On the Shaw Cable side, you could sniff the network all you want, you don't get anything that wasn't explicitly sent to you. On the Videon Cable side, each neighbourhood was wide open -- you could see other people's traffic, file shares, and broadcast packets. Then again, Videon was the same people who left their important systems with the SNMP community set to "public", and were shocked when some mischievous folks decided to fiddle around with them.
You've obviously read Steve Gibson's the-sky-is-falling-because-of-xp's-raw-sockets rant on grc.com. Basically, it makes it easier to spoof, or dump invalid datagrams onto the network. However, assuming all ISPs use proper ingress and egress filtering, and all computers run TCP/IP stacks with secure sequence number generation, the only protocol that would be easy to spoof would be something like UDP, and only if the designers of the specific UDP service forgot to include some kind of secure sequence number in their packets.
So many people who use this argument are overlooking the obvious. An email virus requires an email program that will automatically execute attachments.
Not exactly. All a virus requires is that an e-mail program automatically execute arbitrary code from the e-mail message. Now, this could be done by scripting flaws, or some brain damaged HTML parser. Or, it could by done by a buffer overflow vulnerability in the software that allows an attacker to insert something onto the stack, and have the code jump to his viral code and start executing it. Once this happens, steps 1, 2 and 3 have already been automated by the virus, and you're infected.
Now, I don't care to debate if IE's html engine is more or less buggy than Gecko, or khtml, and frankly, it doesn't matter. A single vulnerable entry point is all an attacker needs to get access.
I would assume they would do it by using two different channels for broadcasting B and G. If I remember properly, it's possible to run 3 different 802.11b networks in a given area without overlapping frequencies. I'm not sure how G works, though. It would make them more expensive, and you'd probably never get full G speeds when a B device was present, but it'd still be faster than B alone.
Re:Lame Canadian radio is based mostly on gov't re
on
Time to Face the Music
·
· Score: 2, Interesting
Except that Radio in the US sucks as much as it does in Canada. Clear Channel, for instance has all the same problems as Canadian stations. Canadian stations have the same formula radio as the US does, except for the requirement for 35% of Canadian content on the radio means a slightly higher emphasis on Canadian music. The fact of the matter is, the music and radio industry have dug their own graves, and it's up to them to demonstrate they shouldn't be laid to rest in them.
Colour laser printers aren't too bad these days. For about $900US, you can get something like the HP LaserJet 2500. The only bad things is the drums in colour printers wear down much faster than monochrome ones, so after 30,000 pages or so, you've got to shell out a lot of money to replace them.
RedHat could do a Linux Directory fairly easily. Most of the low level servers for such a service already exist. OpenLDAP exists and is used by many people, and can plug quite nicely into Samba. The missing parts are mostly convenience related -- admin tools, set-up wizards, etc.
I had a similar experience submitting a bug report... My laptop had a trident video chip in it, and Xv did not work properly on it. The video was offset a different number of pixels depending which output was selected, and overlays larger than 384 pixels wide would not display properly. Alan Hourihane eventually found a way to fix the offset problem (and fixed it again, by adding config options to allow manual adjustment), but couldn't reproduce the 384 pixel overlay problem. As it turned out, what I was trying to say, and what he thought I was saying were two different things. He thought I meant scaling the overlay to >384 pixels, whereas I was meant there was a problem with scaling sources >384 pixels (I guess I wasn't clear enough in my messages).
Eventually I found the problem myself, and it turned out there was just one bit of one Trident register that needed to be changed for the Cyber9385 and higher chipsets to fix the problem. I posted the fix to the xperts mailing list, and he tried to commit a cleaned up version to CVS. Unfortunately, he cleaned it a little too much, and he inadvertently directed it at a completely different set of registers.
I ended up delving far deeper into that world than I ever intended on doing, and found out that the task they perform is far more difficult than most people care to give them credit for. Getting documentation for video chipsets is either difficult or impossible (never did find the documentation for the Cyber9525DVD, which is what my laptop had, only an earlier revision of the chipset), and then learning the layout and internal functionality of XFree86 (it's not undocumented, but it's not entirely documented, either. Documentation is lagged behind implementation by a fair bit). To make things worse, chipset documentation doesn't always tell you what a register does in very clear terms. The culprit in this case was a register that had something to do with line buffers, and to display wide sources, you had to enable 2 line buffers for the overlay (which disables the second overlay).
All in all, it was a minor fix for a minor problem, but it affected me personally, so I did something about it. In this limited experience, it seems that correcting a misinterpretation (regardless of if it's your fault or not) by a developer of your bug report is a rather difficult process.
And, now that the anti-trust suit is over, there's no need for Microsoft to prove that they have any competitors. So, now they can dump Corel, and let them die their inevitable death.
Maybe some day, the opinions you hold will be unpopular, and then who will fight for your rights? When testing a new law, prosecutors go for the most "evil" person they can find, to improve their chances. If they win, they end up establishing precedent, which they can then use against lesser offenders. This means, the ACLU has no choice but to represent the slime balls of the world in trying to defend the rights of everyone else. I seriously doubt that every member of the ACLU believes in every fight they involve themselves in.
There's a reason why virtually every country enumerates the rights of individuals in a document that is near impossible for the government to change on a whim. Criminals should be held responsible for their crimes, no one disputes this. However, a wrongful conviction is must worse than a wrongful dismissal in most cases. In many ways, organizations like the ACLU MUST exist for the legal system to remain healthy. They're a lot like labour unions. They have created a lot of harm, but also created a lot of good. Most workplace safety and worker rights laws are because of their hard work, along with minimum wage, coffee breaks and lunch breaks, etc. They're also somewhat responsible for employee laziness and complacency, and the worst evil of them all -- seniority. Does that mean that labour unions should be abolished? Hell no. They've done, and continue to do important work, and while I may not agree with them a lot of the time, their continued existance is very important to healthy employee/employer relations.
TrollTech is not owned by Canopy Group. Canopy Group has merely invested in them. So has Borland (8.3%), Teknoinvest (3.3%), Orkla (3.3%), and Northzone Ventures (3.3%), for that matter. TrollTech's employees apparently hold 71% of the stock for the company, so some of that could be held by Canopy Group members that happen to be employed by Troll, but it's hard to say the same can't true about Borland, or any of the other investors.
Interesting. Even if you assumed Canopy had a voice within TrollTech, looking at those figures, Inprise/Borland/wahtever-their-name-is-tomorrow has a significantly bigger interest in them (8.3%). I assume they probably picked up troll when the decision was made to use Qt for their Kylix product. The rest of the investors seem to be venture capitalists, just like the Canopy Group.
Why does that reality have to include child pornography, rape, torture, and other mind pollutants?
All those actions you list are already illegal, because they cause quantifable harm to individuals. Those people who are the victims of child pornography, rape and torture often end up scarred for life. However, don't think for a moment that this is about blocking those things.
Do you know what evil is?
The question is, do you know what I think evil is? For a long time rock music was called evil, and repressed by communities all over. These days a lot of people feel the same about the various forms of rap. Do you know what the censorware creators think evil is? And do you think they even care? Do you think the censorware manufacturers give a rats ass what evil is? Not a chance. They only create blacklists for the purpose of making money, and if something gets in the way of them making money, it can end up on the blacklist, too. Do you think breast cancer information is evil? Some censorware makers do. Do you think Amnesty International is evil? How about AIDS information? Sextants? Cucumbers?
Now, if the libraries want to exercise editorial control over the internet access, there's better ways than using blacklist filtering software. They have to do the opposite. Whitelist sites that are approved by the library staff. Of course, you know, just as well as I do, that this would be infeasible for a cash strapped library to do.
Our children are our future, so let's teach them right. Don't listen to the ACLU - they're hell bent on undoing anything that is good and pure in America.
Part of freedom is accepting that people who have ideas that are distasteful have just as much right to speak about them as you do. Just because they're hatefilled, pathetic excuses for human beings doesn't mean they should be repressed. Often times, it's easier to fight these kind of people when they're out in the open, than when they're forced underground, and given false credibility by a government who wishes to stamp them out.
Except you missed a lot of what the BIOS performs today. Today, it's still used to boot the system from various devices (floppy, CD, USB device, hard drive, network), to (hopefully) optimally preconfigure the hardware before the OS looks at it, to provide 32-bit functions for the OS to enumerate PCI devices, to provide APM and ACPI configuration and power management functions. The role of the BIOS hasn't decreased as the years have gone past, but increased.
Regardless of what you want to call it, something has to handle the hardware until the OS can get enough information to intelligently start itself up. That means rudimentary disk I/O (int 13h), video I/O (int 10h), and so on.
The Microsoft knowledge base is getting consistently worse and worse. These days searches turn up practically nothing, and it's getting to be more feasible to search the Microsoft knowledgebase with google, than Microsoft's own search utilities. There's some serious deficiencies with Microsoft's OS's. In particular, the OS never even tries to tell you where a problem occurred. Back in the NT4 days, you'd get a backtrace, information on which system driver caused the fault, and where. It was tremendously useful for diagnosing bad drivers or bad hardware. These days with Windows2000 and XP, it'll give you the exception code (0xc0000005 or something) and four non-descript hexadecimal numbers, and three paragraphs of text apologising for the inconvenience.
Even normal errors have become less descriptive. NT4 used to give error messages like "PROGRAM.EXE: Application Error
This instruction at "0xdeadc0de" referenced memory at "0xdeadbeef". The memory could not be "read"." The wording could've been improved, but it told you what was wrong, and give you the terms you'd need to search for. Ever try to pick out the address/module from a crashing XP application? Either install a debugger, or forget it!
If my computer is going to crash application, so be it. I'd much rather be given a description of what went wrong, than canned false apologies. Knowing that nv4_disp.dll caused my system to crash gives me a lot better place to start fixing things than just "A critical error has occured. We apologize for the inconvenience. [Close]"
I don't know about anyone else, but I feel far safer by the fact that Google is a private company rather than a publically traded company. A private company can afford morals as long as it's profitable. A publically traded company rarely can, because the board will usually pursue increased profits at any and all costs.
It's high time to start regulating the flow of information from the Material Safety Data Sheets. A great deal of them deal with chemical reactions that coule be useful for someone who wishes to cause harm to other peace loving individuals. The obvious solution is to remove the reactivity and safety information from all MSDS sheets, as well as chemical composition information. If we do this, we'll all be safer from the threat of terrorism. It's obvious that the MSDS information poses to benefit to the average person, so can therefore, be edited to remove such harmful portions.
This isn't always an option. Some laptops, like the Toshiba Satellite 5200, that I recently purchased, require ACPI for PCI interupt routing, and this isn't an option that can be loaded as a module. It's either on, or off, or disabled via a kernel parameter. Without ACPI, my sound card, and modem are not assigned IRQs (AMR/CMR devices). To make things worse, the only BIOS setup must be done through Windows XP's control panel, so wiping XP isn't even an option. (Although DVD-R burning and mastering in Linux isn't exactly 'there' yet, so XP has to be used yet again.)
C&C Generals ends up playing a lot more like Warcraft than it does C&C. You have to click on a infantry builder to make infantry, vehicle builder to make vehicles, etc. There are upgrades you can build for units, and there are no more engineers -- infantry takes over buildings. And you build your buildings with a worker unit, instead of from your contruction yard.
To top everything off, single player missions are boring, and repetitive. There's no secrets to find on the maps, or during the game. Compared to Red Alert 2, this game is very lacklustre. RA2 at least had video to tell the story, and the occasional unusual mission (My favorite being with Tania and the three spies, where you have to destroy the nukes. There's several ways to do it.) The "evil" GLA is so predictable, they're no fun to play, versus the "evil" Russians in RA2, which were rather comical.
Slashdot Mirror
See? No paranoia required.
However, I do have some problems with the article. In several of them, it is mentioned that the vertex shader was replaced with a similar looking, but non-identical. Why was NVIDIA able to replace it with a similar but non-identical copy and boost performance so much? Is the 3DMark03 one too biased toward ATI cards, or are the ATI cards too biased towards 3DMark03? I believe these parts alone prove that you can try to make a benchmark try to reveal "real world" performance, but so long as there's multiple ways of doing the same thing without adversely affecting quality or appearance, it's difficult to believe any benchmark, game based or otherwise, is indicitive of real world performance.
There's also the issue of the fact that when those copyrights actually expire, will they be removable from the encryption? When the copyright term expires, they're supposed to be in the public domain, but if you're not legally allowed to break the encryption holding the work, how does it enter the public domain? We've created perpetual copyright without ever really intending to.
Now, if Microsoft wants to provide full, complete and accurate documentation on SMB/CIFS and the MS-RPC function calls, maybe these numbers could be coaxed closer together... But I'm not holding my breath.
The points in that page are quite valid, however, not all configurations are necessarily vulnerable to an ARP poisoning attack. For example, many cable modems are capable of filtering which traffic is meant for them and which is not. In order to do ARP spoofing, you need to hear an ARP request to respond to, so if you never receive a request, you can't return a response. However, this isn't the case for all cable networks. In Winnipeg, we had both. On the Shaw Cable side, you could sniff the network all you want, you don't get anything that wasn't explicitly sent to you. On the Videon Cable side, each neighbourhood was wide open -- you could see other people's traffic, file shares, and broadcast packets. Then again, Videon was the same people who left their important systems with the SNMP community set to "public", and were shocked when some mischievous folks decided to fiddle around with them.
You've obviously read Steve Gibson's the-sky-is-falling-because-of-xp's-raw-sockets rant on grc.com. Basically, it makes it easier to spoof, or dump invalid datagrams onto the network. However, assuming all ISPs use proper ingress and egress filtering, and all computers run TCP/IP stacks with secure sequence number generation, the only protocol that would be easy to spoof would be something like UDP, and only if the designers of the specific UDP service forgot to include some kind of secure sequence number in their packets.
Now, I don't care to debate if IE's html engine is more or less buggy than Gecko, or khtml, and frankly, it doesn't matter. A single vulnerable entry point is all an attacker needs to get access.
I would assume they would do it by using two different channels for broadcasting B and G. If I remember properly, it's possible to run 3 different 802.11b networks in a given area without overlapping frequencies. I'm not sure how G works, though. It would make them more expensive, and you'd probably never get full G speeds when a B device was present, but it'd still be faster than B alone.
Except that Radio in the US sucks as much as it does in Canada. Clear Channel, for instance has all the same problems as Canadian stations. Canadian stations have the same formula radio as the US does, except for the requirement for 35% of Canadian content on the radio means a slightly higher emphasis on Canadian music. The fact of the matter is, the music and radio industry have dug their own graves, and it's up to them to demonstrate they shouldn't be laid to rest in them.
Colour laser printers aren't too bad these days. For about $900US, you can get something like the HP LaserJet 2500. The only bad things is the drums in colour printers wear down much faster than monochrome ones, so after 30,000 pages or so, you've got to shell out a lot of money to replace them.
Except, this mod involves using the NTSC output on the video card, which means things will be generally unreadable on the screen.
RedHat could do a Linux Directory fairly easily. Most of the low level servers for such a service already exist. OpenLDAP exists and is used by many people, and can plug quite nicely into Samba. The missing parts are mostly convenience related -- admin tools, set-up wizards, etc.
They're pretty much the only video tuner manufacturer out there. I don't think the chipset has evolved much since Conexant bought them, though.
Eventually I found the problem myself, and it turned out there was just one bit of one Trident register that needed to be changed for the Cyber9385 and higher chipsets to fix the problem. I posted the fix to the xperts mailing list, and he tried to commit a cleaned up version to CVS. Unfortunately, he cleaned it a little too much, and he inadvertently directed it at a completely different set of registers.
I ended up delving far deeper into that world than I ever intended on doing, and found out that the task they perform is far more difficult than most people care to give them credit for. Getting documentation for video chipsets is either difficult or impossible (never did find the documentation for the Cyber9525DVD, which is what my laptop had, only an earlier revision of the chipset), and then learning the layout and internal functionality of XFree86 (it's not undocumented, but it's not entirely documented, either. Documentation is lagged behind implementation by a fair bit). To make things worse, chipset documentation doesn't always tell you what a register does in very clear terms. The culprit in this case was a register that had something to do with line buffers, and to display wide sources, you had to enable 2 line buffers for the overlay (which disables the second overlay).
All in all, it was a minor fix for a minor problem, but it affected me personally, so I did something about it. In this limited experience, it seems that correcting a misinterpretation (regardless of if it's your fault or not) by a developer of your bug report is a rather difficult process.
And, now that the anti-trust suit is over, there's no need for Microsoft to prove that they have any competitors. So, now they can dump Corel, and let them die their inevitable death.
There's a reason why virtually every country enumerates the rights of individuals in a document that is near impossible for the government to change on a whim. Criminals should be held responsible for their crimes, no one disputes this. However, a wrongful conviction is must worse than a wrongful dismissal in most cases. In many ways, organizations like the ACLU MUST exist for the legal system to remain healthy. They're a lot like labour unions. They have created a lot of harm, but also created a lot of good. Most workplace safety and worker rights laws are because of their hard work, along with minimum wage, coffee breaks and lunch breaks, etc. They're also somewhat responsible for employee laziness and complacency, and the worst evil of them all -- seniority. Does that mean that labour unions should be abolished? Hell no. They've done, and continue to do important work, and while I may not agree with them a lot of the time, their continued existance is very important to healthy employee/employer relations.
TrollTech is not owned by Canopy Group. Canopy Group has merely invested in them. So has Borland (8.3%), Teknoinvest (3.3%), Orkla (3.3%), and Northzone Ventures (3.3%), for that matter. TrollTech's employees apparently hold 71% of the stock for the company, so some of that could be held by Canopy Group members that happen to be employed by Troll, but it's hard to say the same can't true about Borland, or any of the other investors.
Interesting. Even if you assumed Canopy had a voice within TrollTech, looking at those figures, Inprise/Borland/wahtever-their-name-is-tomorrow has a significantly bigger interest in them (8.3%). I assume they probably picked up troll when the decision was made to use Qt for their Kylix product. The rest of the investors seem to be venture capitalists, just like the Canopy Group.
Now, if the libraries want to exercise editorial control over the internet access, there's better ways than using blacklist filtering software. They have to do the opposite. Whitelist sites that are approved by the library staff. Of course, you know, just as well as I do, that this would be infeasible for a cash strapped library to do.
Part of freedom is accepting that people who have ideas that are distasteful have just as much right to speak about them as you do. Just because they're hatefilled, pathetic excuses for human beings doesn't mean they should be repressed. Often times, it's easier to fight these kind of people when they're out in the open, than when they're forced underground, and given false credibility by a government who wishes to stamp them out.Regardless of what you want to call it, something has to handle the hardware until the OS can get enough information to intelligently start itself up. That means rudimentary disk I/O (int 13h), video I/O (int 10h), and so on.
Even normal errors have become less descriptive. NT4 used to give error messages like "PROGRAM.EXE: Application Error This instruction at "0xdeadc0de" referenced memory at "0xdeadbeef". The memory could not be "read"." The wording could've been improved, but it told you what was wrong, and give you the terms you'd need to search for. Ever try to pick out the address/module from a crashing XP application? Either install a debugger, or forget it!
If my computer is going to crash application, so be it. I'd much rather be given a description of what went wrong, than canned false apologies. Knowing that nv4_disp.dll caused my system to crash gives me a lot better place to start fixing things than just "A critical error has occured. We apologize for the inconvenience. [Close]"
I don't know about anyone else, but I feel far safer by the fact that Google is a private company rather than a publically traded company. A private company can afford morals as long as it's profitable. A publically traded company rarely can, because the board will usually pursue increased profits at any and all costs.
It's high time to start regulating the flow of information from the Material Safety Data Sheets. A great deal of them deal with chemical reactions that coule be useful for someone who wishes to cause harm to other peace loving individuals. The obvious solution is to remove the reactivity and safety information from all MSDS sheets, as well as chemical composition information. If we do this, we'll all be safer from the threat of terrorism. It's obvious that the MSDS information poses to benefit to the average person, so can therefore, be edited to remove such harmful portions.
This isn't always an option. Some laptops, like the Toshiba Satellite 5200, that I recently purchased, require ACPI for PCI interupt routing, and this isn't an option that can be loaded as a module. It's either on, or off, or disabled via a kernel parameter. Without ACPI, my sound card, and modem are not assigned IRQs (AMR/CMR devices). To make things worse, the only BIOS setup must be done through Windows XP's control panel, so wiping XP isn't even an option. (Although DVD-R burning and mastering in Linux isn't exactly 'there' yet, so XP has to be used yet again.)
To top everything off, single player missions are boring, and repetitive. There's no secrets to find on the maps, or during the game. Compared to Red Alert 2, this game is very lacklustre. RA2 at least had video to tell the story, and the occasional unusual mission (My favorite being with Tania and the three spies, where you have to destroy the nukes. There's several ways to do it.) The "evil" GLA is so predictable, they're no fun to play, versus the "evil" Russians in RA2, which were rather comical.