There are MANY health risks of COX-2 inhibitors which have only been recently discovered. The dangers are not cardiac (pertaining to the heart) but cardiovascular, COX-2 is an enzyme which regulates a number of cardiovascular functions, most notably the inflamatory response, but also the formation of new blood vessels following injury, and your body's natural defenses against blood clots flowing through your circulatory system. The latter is what increases the danger of heart attack or stroke, both of which can be caused by blood clots clogging your cardiac arteries or brain arteries respectively.
A report just released in the December issue of Nature Medicine claims that blocking the COX-2 enzyme interferes with angiogenesis (the formation of new blood vessels). New blood vessels are imperative to wound and ulcer healing. Cited laboratory study results revealed reduced angiogenesis when rat or human blood cells were treated with either indomethacin (an older NSAID) or by NS-398 (a COX-2 inhibitor). COX-2 and COX-1 are required for healing of the stomach and intestinal lining. This new data raises questions concerning whether selective COX-2 drugs are safer for the gastrointestinal tract than older NSAIDS after all since inhibiting angiogenesis may result in ulcer complications.
FDA is issuing an advisory because of recently released data from controlled clinical trials showing that the Cox-2 selective agents (Vioxx, Celebrex and Bextra) may be associated with an increased risk of serious cardiovascular events (heart attack and stroke) especially when they are used for long periods of time or in very high-risk settings (immediately after heart surgery), as the COX-2 enzyme apperas to play a key role in the body's natural ability to combat blood clots," the FDA said in its press release.
That said, the majority of pharmeceudical spam I receive is for Cialias and Viagra, the anti-impotency pills. Does anyone here actually receive spam for Vioxx more frequently than these?
...than "Who Wants to Be a Millionaire?"
Finally, we'll see people winning millions by answering... "answers" like "These particles are either baryonic or fermionic" rather than bullshit questions like "What is the capital of California? A) Los Angeles B) San Diego C) Sacremento D) Dollywood"
Okay, a number of things here...
The entire iBook and PowerBook lines are both G4. This isn't a good position for Apple to be in because it lowers the differentiation between the two. It's inevitable that Apple will eventually get the PPC970FX (which includes power management needed for laptops, currently used in the G5 Xserve) into the PowerBook line, or some other PPC970 derivative, at which point, at which point it's likely they will [i]not[/i] do the same for the iBook line, keeping it G4-based and reserving the G5 for the PowerBook line exclusively. So, Apple will [i]not[/b] be dumping the G4 until they are ready to upgrade the iBook line, and given how long it was since the G4 came out for them to release a G4 iBook (roughly 5 years) it's going to be quite awhile before they do that.
Secondly, Motorola is going to spin off their chip division as Freescale Semiconductor. They've recently released a new G4 rev with a number of new features included an upgraded bus speed.
Finally, IBM has a number of high performance G3 PowerPC chips to which the PPC970's AltiVec unit could be added, such as the 750CX, making a G4-like processor which could be used to replace the current Motorola G4s.
Viruses are now including multiple attack vectors, and often times some of these require human intervention while some don't. As viruses grow increasingly multiparadigm and begin exhibiting both the properties of the canonical virus (requiring human intervention) and worm (spreading without human intervention) the semantic distinction grows less important.
This is a distinction which Joe Sixpack has a terrible time grasping. Telling someone "Your computer's got worms!" is less likely to be comprehend than "Your computer has a virus", further complicating the difficulty of explaining to Joe Sixpack that hardware buffer overflow protection could save him from the next Windows worm...
For those who don't know, the Hyugens probe bound for Titan was developed by the EU's Space Agency. It will provide us with the first glimpses below Titan's cloudy surface, and was carried by NASA's Cassini probe.
It's wonderful to see such collaboration between the ESA and NASA, and I hope we continue to see such efforts in the future.
They're trying to take on Big Blue, one of the largest corporations in the world who also possesses one of the largest and most skilled legal teams.
This was inevitable.
What's the make/model of your laptop and desktop, and what build of Solaris Express did you try? Only recently has it shipped with amd64 support, although it's surprising that it wasn't able to boot in IA32 mode. What error were you encountering?
It seems to me that they have taken some action besides just grandstanding. They have resurrected the x86 version and added several interesting features--containers, DTrace, and ZFS, for example--that are available today in beta versions of Solaris 10. They're actively rounding up support from developers and software companies. And they announced that the production version of Solaris 10 on x86 will be available for free. What do you think about the x86 move and the new Solaris features?
Solaris/x86 is a joke, last I heard. (It has) very little support for any kind of strange hardware. If you thought Linux had issues with driver availability for some things, let's see you try Solaris/x86.
This attitude is perfectly fine for any Solaris release prior to 10. However, Sun has made massive strides both in performance and hardware support, especially on the AMD64 platform.
Furthermore, he completely dodged the questions about containers, DTrace, and ZFS. While these are all fancy names for things which are also available in Linux, the truth of the matter is the Linux counterparts cannot hold a candle to any of these features in Solaris. Here's a quick run down:
Containers: Solaris's new virtualization mechanism. Containers have a special kernel image which is able to communicate with the main system kernel entirely in kernel space. This is somewhat similar to the approach taken by the Xen virtual machine, except that Xen does it at a much lower level. Solaris containers may be thought of as somewhere between a Linux kernel instance running in Xen on top of another Linux kernel and BSD jails. It certainly is nothing like UML, where the UML kernel is running in process context and thus performs rather pathetically.
ZFS: This integrates all the features of a high end filesystem and high end volume manager into a single package. Unfortunately, this will only be available a few months after Solaris 10 General Availability, but once it hits expect tools like VxFS and the Veritas Volume Manager to be rendered thoroughly obsolete on the Solaris side. Linux certainly has many interesting filesystems with cool whiz-bang features, many of which aren't implemented in ZFS, but on the flip side ZFS has many features tuned towards the enterprise market which are seen in very few Linux filesystems, most notably XFS.
DTrace: While a bit obtuse for the time being, a simple demonstration of its power must be seen. The main advantage DTrace has over Linux alternatives such as KProbes, besides being massively more powerful, is that there is no performance impact on the system when they are not in use. DTrace probes are inserted into the kernel when needed and removed when not, whereas KProbes require they statically be built into the kernel.
Conclusion: There is a considerable amount of feature parity between Linux and Solaris 10, but the Solaris features all have an edge over the Linux ones. Linus should not let his hubris cloud his judgement... I expect Solaris 10 to be a major competator to Linux in the low end SMP server market.
Right now running Linux (or FreeBSD) on AMD64 has you flying by the seat of your pants a bit... it's certainly not polished and there are a number of caveats and gotchas to watch out for. Contrarily Solaris 10/AMD64, especially on Sun's own hardware, runs like a dream. I expect Solaris 10 to thoroughly decimate Linux in the Opteron server market.
There are still a number of areas where Linux is still playing catch up to Solaris as well, most notably in the realm of schedulers. While Linux 2.6 now sports a constant time scheduler like Solaris has had for a half decade, Solaris still supports modular schedulers which can be swapped in and out, can be active simultaneously, and processes can be moved between them. One of the most notable ones fo
For a workplace, there's no better solution than single-sign-on (Kerberos or the like) using a SmartCard. People understand how to keep something like a key safe, but keeping a bit of information safe, especially when it's something they have to keep in their head, is considerably more difficult.
I think the best approach is something like a Sun Microsystems Sunray environment where you can stick your SmartCard into any Sunray and instantly pull up your session from the server. Instead of having to "log out" you simply pull your SmartCard out of the Sunray, and that's the end of your session (even though it stays going on the server)
http://beacon.sf.net/ tries to do this using UDP and filesystem monitoring. It waits for the RSS document to change then sends a UDP datagram to notify everyone that a new version is available. It's better than everyone polling the server via HTTP anyway.
Seems to me like BitTorrent's days are numbered, it's halfway between two other technologies and doesn't fit the role of either of them. This application seems to be merging BitTorrent with what Kazaa originally tried to be, a fully decentralized distributed network that allowed parallel downloads of file chunks. Now given BitTorrent does parallel distributed downloads of file chunks substantially better than Kazaa, and it sounds like this new application will fill the P2P role of BitTorrent much better.
On the centralized aspect, I think when it's complete PDTP will completely replace BitTorrent for downloads from a central server. It can easily be used with large file repositories like current FTP servers maintain. It works way more like Apache, you just point it at the directory you want it to serve and it does the rest, no need to generate crappy.torrent files. Plus since all of the logic is server side it makes writing clients much easier, and allows for substantially faster downloads and better control. I always get sick of when my torrents are coming in 5kpbs down 100kbps up:(
Seems especially pointless considering it has already been accomplished about 10 years ago by Wacom...
Comparing UML to N1 Grid Containers? Ridiculous..
on
Sun-isms Debunked
·
· Score: 4, Informative
GNU/Linux may not have Solaris containers (which allow applications to run in virtual instances of Solaris, isolated from the rest of the OS), but it does have Usermode Linux (UML) which provides similar functionality using a different technique.
UML has substantially low performance compared to N1 Grid Containers. If you're going to compare a server virtualization feature, compare to something like the Xen Virtual Machine, in this performance comparison, you can see the performance of UML is rather appalling, especially compared to Xen.
The performance of Solaris Grid Containers is more akin to Xen or FreeBSD jails. However, the advantage N1 Grid Containers have over Xen is that they are portable to every platform Solaris runs on (SPARC, IA32, AMD64) whereas Xen only emulates one platform (IA32). Also, other Solaris features to which there are currently no Linux counterparts such as the Fair Share Scheduler, which allows a N1 Grid Container to be bound to certain processors, and given a dedicated percentage (or share) of available processor resources. This provides an advantage over Xen and UML which can't even use multiple CPUs. It has an advantage over FreeBSD jails where monopolization of system resources by a single jail cannot be easily avoided.
While Linux may have counterparts to various Solaris features, in terms of maturity, feature set, and performance of these features Solaris has Linux trumped.
Wrong. The name Unix is an abbreviation of UNICS, which stands for UNiplexed Information and Computing Service, a play on MULTICS, MULTiplexed Information and Computing Service
"the total economic impact of the Minnesota ethanol industry was estimated at $588 million in 2002. In comparison, the state's ethanol subsidy for the year was $33.7 million that means the economic impact was 17 times the subsidy payment."
And remember, you're talking about ethanol as opposed to gasoline, which we get from terrorist nations, which costs over twice as much as E85 fuels (E85 sells for $0.90/gallon) and pollutes substantially more.
It takes more energy to make ethanol than you get out of it.
...except through the use of reformers which extract hydrogen from gasoline or ethanol which can be in turn used in conjunction with hydrogen fuel cells.
I have much better hopes for E85 fuel, which combines 85% ethanol and 15% gasoline. Any existing car can be modified to run on E85 in addition to regular gasoline, meaning that people don't need to throw away their existing vehicles and buy a new hydrogen car. Since they can still run on gasoline, you don't need to worry about looking for a specific kind of fuel station... buy E85 if it's available and regular gasoline if it isn't.
E85 is also substantially more environmentally friendly than gasoline:
E85 is environmentally-friendly. It has the highest oxygen content of any fuel available today, making it burn more completely (cleaner) than gasoline. E85 contains 80% less gum-forming compounds, like the olefins found in gasoline. Production and use of E85 results in a nearly 30% reduction in greenhouse gas emissions. More than 100 major U.S. cities suffer from unhealthy levels of smog. E85 may be able to help. U.S. Environmental Protection Agency (EPA) studies have shown that high-blend ethanol fuels can reduce harmful exhaust emissions by more than 50 percent and smog-forming pollution by 15% or more.
E85 can be produced from surplus feed corn which is otherwise thrown out every year. Our nation has the capacity to manufacture it in quantity, whereas hydrogen is currently difficult to produce and expensive, and the easiest chemical processes by which to produce it result in toxic chemicals (such as reacting metal with hydroxides). Furthermore, everything in our current gasoline infractructure, from tanker trunks, storage tanks, gas pumps, and vehicles, can be used with E85, whereas hydrogen would require that we retool our entire infrastructure.
E85 would also open the doors to a new class of pure ethanol vehicles, including, as I said earlier, ones which use reformers to extract hydrogen from ethanol and run it through fuel cells, virtually eliminating pollution and the inefficiency of internal combustion engines. E85 would move our source of energy from terrorist controlled oil to domestically produced corn and other starchy crops. E85 would allow us to utilize surplus starchy crops rather than just throwing them away, eliminating waste.
All in all, I don't see what the buzz about hydrogen is all about. It would require an impractical infrastructural transition which is unlikely to happen until we've thoroughly exhausted our oil supplies. E85 lets us keep our whole existing infrastructure while still solving most of the problems attributed to oil.
Things like the "no-execute" bit on modern CPUs provide an additional layer of security and prevent purposely damaged code from running directly on the CPU. However, until operating systems implement this in their own code, any application that does not want to adhere to the no-exec flag does not have to. This is like flock on Unix which only sets a file locking flag which applications are expected to obey rather than true file locking as implemented on other systems.
Wrong. sparcv9, for example, implements a non-executable user stack per default. In POSIX, all memory from the heap is pre-marked non-executable (on architectures that support page protections) unless it is explicitly set by the program to be executable (for example, in JIT compilers) using functions like mprotect(). In Windows, this is implemented as a flag passed to HeapAlloc().
The interface design and OS support is already there, what isn't is people buying non-IA32 CPUs in large numbers.
Can't beat my web site('s name): bush.fails.org
on
Pre-Election Discussion
·
· Score: 2, Informative
Complete coverage of how Bush has failed to prevent A.Q. Khan, the world's top nuclear arms dealer, from selling the recipe and ingredients to nuclear bombs to countless rogue nations and terrorist groups, like North Korea, Iran, Libya, and Al Qaida, including the complete history of how Pakistan got the bomb in the first place (Hint: they bought the design from China)
If you thin Bush has done a good job protecting America's national security this is an eye-opening must read.
Even in the enterprise and scientific community cost is still a consideration. I come from a scientific computing group that used to pay $20,000 for dual UltraSPARC-III Sun Blade workstations, then moved to Linux clusters, and is now moving to Apple clusters and Apple workstations. After being given a large discount by Apple, we found Apple's offer of cluster of 20 dual 2.0GHz G5 Xserves to be more powerful (and certainly less problematic) than Dell's offer of 20 3.2GHz Nocona Xeons. We found the new Intel Fortran Compiler 8.1 with EM64T support to be rather underwelming... its binaries optimized for the 64-bit Nocona Xeons with SSE3 couldn't outperform those made by Pathscale, the leading AMD64 compiler suite, on Intel's own processors (even though Pathscale only supports SSE2). However, neither could outperform IBM XL Fortran on the Xserve's 2.0GHz IBM PPC970FX processors with AltiVec units.
At less than an eighth of the price of a Sun workstation, you can purchase a dual 2.5GHz G5, which lacks many of the amenities of Sun Blades such as ECC RAM and 10,000RPM FC-AL hard drives, the model runs considerably faster at a fraction of the price, and the system can double as a user desktop with both Unix (i.e. scientific computing programs) and (otherwise) Windows amenities such as Microsoft Office and Adobe tools (Photoshop/Illustrator/Acrobat).
For any role I can imagine for a dual Opteron workstation, I can see a G5 in the same role for a considerably cheaper price. Furthermore, I can see a G5 outperforming an Opteron in any of those roles, because in virtually all of them (scientific computing, medical computing, multimedia/3D modelling/video production) the AltiVec unit on the G5 will be extremely beneficial, whereas Opteron has no good vector units for these purposes (Opteron SSE2 is slower than its FPU, SSE is only 64-bits, doesn't support double precision floating point or the multitude of operations AltiVec supports such as trig functions needed for FFT/DCT transforms)
I believe that next to the new Nocona Xeon-based Dell Precision workstations (with SSE3 which is comparable to AltiVec), Apple has the cheapest and most powerful Tier 1 workstation offering in the form of the dual 2.5GHz G5, at least for the roles a high end dual processor 64-bit workstation is intended to serve.
Especially if we're talking Longhorn...
on
How Cheap Can A PC Be?
·
· Score: 2, Interesting
For $100, I really doubt you're going to be able to get Longhorn-caliber hardware for quite some time, considering Longhorn's system requirements call for:
A dual-core CPU running at 4 to 6GHz
A minimum of 2 gigs of RAM
A terabyte of storage
A graphics processor that runs three times faster than those on the market today
Subpixel rendering has been around for quite a long time. Two things that I can think of right off the bat are Microsoft's ClearType and FreeType, both of which have hinting engines which support subpixel rendering.
Subpixel rendering takes into account the physical position of the red, green, and blue subpixels of an LCD display, and can therefore provide up to 3X the horizontal resolution of a typical display (with distortion, of course)
Slashdot Mirror
Here's some background information:
http://arthritis.about.com/od/cox2inhibitors/a/saf etyreport.htm
http://money.cnn.com/services/tickerheadlines/djh/ 200412231648DOWJONESDJONLINE000879.htm
That said, the majority of pharmeceudical spam I receive is for Cialias and Viagra, the anti-impotency pills. Does anyone here actually receive spam for Vioxx more frequently than these?
...than "Who Wants to Be a Millionaire?" Finally, we'll see people winning millions by answering... "answers" like "These particles are either baryonic or fermionic" rather than bullshit questions like "What is the capital of California? A) Los Angeles B) San Diego C) Sacremento D) Dollywood"
Secondly, Motorola is going to spin off their chip division as Freescale Semiconductor. They've recently released a new G4 rev with a number of new features included an upgraded bus speed.
Finally, IBM has a number of high performance G3 PowerPC chips to which the PPC970's AltiVec unit could be added, such as the 750CX, making a G4-like processor which could be used to replace the current Motorola G4s.
This is a distinction which Joe Sixpack has a terrible time grasping. Telling someone "Your computer's got worms!" is less likely to be comprehend than "Your computer has a virus", further complicating the difficulty of explaining to Joe Sixpack that hardware buffer overflow protection could save him from the next Windows worm...
http://planetary.org/saturn/images_spacecraft.html
It's wonderful to see such collaboration between the ESA and NASA, and I hope we continue to see such efforts in the future.
They're trying to take on Big Blue, one of the largest corporations in the world who also possesses one of the largest and most skilled legal teams. This was inevitable.
What's the make/model of your laptop and desktop, and what build of Solaris Express did you try? Only recently has it shipped with amd64 support, although it's surprising that it wasn't able to boot in IA32 mode. What error were you encountering?
This attitude is perfectly fine for any Solaris release prior to 10. However, Sun has made massive strides both in performance and hardware support, especially on the AMD64 platform.
Furthermore, he completely dodged the questions about containers, DTrace, and ZFS. While these are all fancy names for things which are also available in Linux, the truth of the matter is the Linux counterparts cannot hold a candle to any of these features in Solaris. Here's a quick run down:
Containers: Solaris's new virtualization mechanism. Containers have a special kernel image which is able to communicate with the main system kernel entirely in kernel space. This is somewhat similar to the approach taken by the Xen virtual machine, except that Xen does it at a much lower level. Solaris containers may be thought of as somewhere between a Linux kernel instance running in Xen on top of another Linux kernel and BSD jails. It certainly is nothing like UML, where the UML kernel is running in process context and thus performs rather pathetically.
ZFS: This integrates all the features of a high end filesystem and high end volume manager into a single package. Unfortunately, this will only be available a few months after Solaris 10 General Availability, but once it hits expect tools like VxFS and the Veritas Volume Manager to be rendered thoroughly obsolete on the Solaris side. Linux certainly has many interesting filesystems with cool whiz-bang features, many of which aren't implemented in ZFS, but on the flip side ZFS has many features tuned towards the enterprise market which are seen in very few Linux filesystems, most notably XFS.
DTrace: While a bit obtuse for the time being, a simple demonstration of its power must be seen. The main advantage DTrace has over Linux alternatives such as KProbes, besides being massively more powerful, is that there is no performance impact on the system when they are not in use. DTrace probes are inserted into the kernel when needed and removed when not, whereas KProbes require they statically be built into the kernel.
Conclusion: There is a considerable amount of feature parity between Linux and Solaris 10, but the Solaris features all have an edge over the Linux ones. Linus should not let his hubris cloud his judgement... I expect Solaris 10 to be a major competator to Linux in the low end SMP server market.
Right now running Linux (or FreeBSD) on AMD64 has you flying by the seat of your pants a bit... it's certainly not polished and there are a number of caveats and gotchas to watch out for. Contrarily Solaris 10/AMD64, especially on Sun's own hardware, runs like a dream. I expect Solaris 10 to thoroughly decimate Linux in the Opteron server market.
There are still a number of areas where Linux is still playing catch up to Solaris as well, most notably in the realm of schedulers. While Linux 2.6 now sports a constant time scheduler like Solaris has had for a half decade, Solaris still supports modular schedulers which can be swapped in and out, can be active simultaneously, and processes can be moved between them. One of the most notable ones fo
Diebold bought the election for Bush. Here's a funny comic on the matter: http://www.bushspeaks.com/img/rove-poses-for-diebo ld-ad.jpg
I think the best approach is something like a Sun Microsystems Sunray environment where you can stick your SmartCard into any Sunray and instantly pull up your session from the server. Instead of having to "log out" you simply pull your SmartCard out of the Sunray, and that's the end of your session (even though it stays going on the server)
http://beacon.sf.net/ tries to do this using UDP and filesystem monitoring. It waits for the RSS document to change then sends a UDP datagram to notify everyone that a new version is available. It's better than everyone polling the server via HTTP anyway.
On FreeBSD, it's all about mtree...
On the centralized aspect, I think when it's complete PDTP will completely replace BitTorrent for downloads from a central server. It can easily be used with large file repositories like current FTP servers maintain. It works way more like Apache, you just point it at the directory you want it to serve and it does the rest, no need to generate crappy .torrent files. Plus since all of the logic is server side it makes writing clients much easier, and allows for substantially faster downloads and better control. I always get sick of when my torrents are coming in 5kpbs down 100kbps up :(
Seems especially pointless considering it has already been accomplished about 10 years ago by Wacom...
UML has substantially low performance compared to N1 Grid Containers. If you're going to compare a server virtualization feature, compare to something like the Xen Virtual Machine, in this performance comparison, you can see the performance of UML is rather appalling, especially compared to Xen.
The performance of Solaris Grid Containers is more akin to Xen or FreeBSD jails. However, the advantage N1 Grid Containers have over Xen is that they are portable to every platform Solaris runs on (SPARC, IA32, AMD64) whereas Xen only emulates one platform (IA32). Also, other Solaris features to which there are currently no Linux counterparts such as the Fair Share Scheduler, which allows a N1 Grid Container to be bound to certain processors, and given a dedicated percentage (or share) of available processor resources. This provides an advantage over Xen and UML which can't even use multiple CPUs. It has an advantage over FreeBSD jails where monopolization of system resources by a single jail cannot be easily avoided.
While Linux may have counterparts to various Solaris features, in terms of maturity, feature set, and performance of these features Solaris has Linux trumped.
By 1996 WebObjects was pretty much all NeXT had left after sales of NeXT and NeXT systems had plummeted to vitually nil.
Wrong. The name Unix is an abbreviation of UNICS, which stands for UNiplexed Information and Computing Service, a play on MULTICS, MULTiplexed Information and Computing Service
Wrong. According to this Minnesota Business Journal article:
"the total economic impact of the Minnesota ethanol industry was estimated at $588 million in 2002. In comparison, the state's ethanol subsidy for the year was $33.7 million that means the economic impact was 17 times the subsidy payment."
And remember, you're talking about ethanol as opposed to gasoline, which we get from terrorist nations, which costs over twice as much as E85 fuels (E85 sells for $0.90/gallon) and pollutes substantially more.
Wrong. Even in 1988 energy generated by the ethanol exceeded energy inputs by 16%. Nowadays that number is closer to 34%, according to a USDA study.
I have much better hopes for E85 fuel, which combines 85% ethanol and 15% gasoline. Any existing car can be modified to run on E85 in addition to regular gasoline, meaning that people don't need to throw away their existing vehicles and buy a new hydrogen car. Since they can still run on gasoline, you don't need to worry about looking for a specific kind of fuel station... buy E85 if it's available and regular gasoline if it isn't.
E85 is also substantially more environmentally friendly than gasoline:
E85 can be produced from surplus feed corn which is otherwise thrown out every year. Our nation has the capacity to manufacture it in quantity, whereas hydrogen is currently difficult to produce and expensive, and the easiest chemical processes by which to produce it result in toxic chemicals (such as reacting metal with hydroxides). Furthermore, everything in our current gasoline infractructure, from tanker trunks, storage tanks, gas pumps, and vehicles, can be used with E85, whereas hydrogen would require that we retool our entire infrastructure.
E85 would also open the doors to a new class of pure ethanol vehicles, including, as I said earlier, ones which use reformers to extract hydrogen from ethanol and run it through fuel cells, virtually eliminating pollution and the inefficiency of internal combustion engines. E85 would move our source of energy from terrorist controlled oil to domestically produced corn and other starchy crops. E85 would allow us to utilize surplus starchy crops rather than just throwing them away, eliminating waste.
All in all, I don't see what the buzz about hydrogen is all about. It would require an impractical infrastructural transition which is unlikely to happen until we've thoroughly exhausted our oil supplies. E85 lets us keep our whole existing infrastructure while still solving most of the problems attributed to oil.
Wrong. sparcv9, for example, implements a non-executable user stack per default. In POSIX, all memory from the heap is pre-marked non-executable (on architectures that support page protections) unless it is explicitly set by the program to be executable (for example, in JIT compilers) using functions like mprotect(). In Windows, this is implemented as a flag passed to HeapAlloc().
The interface design and OS support is already there, what isn't is people buying non-IA32 CPUs in large numbers.
If you thin Bush has done a good job protecting America's national security this is an eye-opening must read.
At less than an eighth of the price of a Sun workstation, you can purchase a dual 2.5GHz G5, which lacks many of the amenities of Sun Blades such as ECC RAM and 10,000RPM FC-AL hard drives, the model runs considerably faster at a fraction of the price, and the system can double as a user desktop with both Unix (i.e. scientific computing programs) and (otherwise) Windows amenities such as Microsoft Office and Adobe tools (Photoshop/Illustrator/Acrobat).
For any role I can imagine for a dual Opteron workstation, I can see a G5 in the same role for a considerably cheaper price. Furthermore, I can see a G5 outperforming an Opteron in any of those roles, because in virtually all of them (scientific computing, medical computing, multimedia/3D modelling/video production) the AltiVec unit on the G5 will be extremely beneficial, whereas Opteron has no good vector units for these purposes (Opteron SSE2 is slower than its FPU, SSE is only 64-bits, doesn't support double precision floating point or the multitude of operations AltiVec supports such as trig functions needed for FFT/DCT transforms)
I believe that next to the new Nocona Xeon-based Dell Precision workstations (with SSE3 which is comparable to AltiVec), Apple has the cheapest and most powerful Tier 1 workstation offering in the form of the dual 2.5GHz G5, at least for the roles a high end dual processor 64-bit workstation is intended to serve.
Try getting that for $100
Subpixel rendering takes into account the physical position of the red, green, and blue subpixels of an LCD display, and can therefore provide up to 3X the horizontal resolution of a typical display (with distortion, of course)
Here's a nice writeup