Realistically, it doesn't matter what the Bush administration wants to see done in Iran. Establishing lasting Iraqi democracy will take years, and the volunteer force we have now is being stretched thin by our Iraqi presence. The fundamental strategy of war is to go in with more forces than your opponent. And the more so, the better-- less causualties are expected when you've got ten times as many troops than one extra troop. This isn't Risk.
So we went into iraq on a high casulty mission. Fortunately, this turned out to be a non-issue; the military forces Congress was told Saddam was massing in preperation was non-existant. However, the defense department lowballed the estimated need for troops by about 60 percent. Shinseki was asking for large numbers, to reduce casulties and occupy the region. It's beginning (if not already) to appear that he was at least more correct, as attacks continue to increase, and foreign fighters begin falling into the region.
So maybe we're seeing some casulties, but it's only a small number of the hundred thousand or so deployed, right? Well, those casulties are significant in a volunteer army. Enlisted soldiers are great, but enlistment itself is violently tied to public opinion for current events and leadership, even though service commitments are longer than any elected position lasts. In a vicious cycle, that negative impact has negative reprocussions-- stop loss measures enacted to keep enlistment high. This is basically a form of drafting people highly skilled in military needs (ie former soldiers). As this becomes public news, it further erodes the desire to enlist amongst the general population.
At this point, invading Iran would require acts of severe delusion by Congressmembers, or a draft (assuming those can be distinct possibilities). The draft is the missing element to a Vietnam re-enactment in 2005. I don't suspect anyone in Congress is ready to face that sort of pressure.
The first time I added the threads tab, I was shocked that so many programs actually included threading, given how little time we spent on threading in my classes, and how much the proffessors indicated that threading was a difficult problem to solve.
I'm not sure where the extra threads come from in normal applications; perhaps there's some threading involved with the basic UI. For example, it appears that spider.exe creates a new thread to animate a new deal of cards.
At any rate, the parent is correct; games are multithreaded, I even admit this in my parent post. But they're not exactly optimizable for threads, which was my point, and which the grand-grand parent explicitly admitted. But if all the work happens in a single thread, then multiprocessing can't ever really help.
You're already somewhat ahead of the game. However, generally speaking, you've got one, maybe two intensive threads.
And your threading for a FPS is going to be highly interdependent. The process is generally: 1. get new packet 2. get input 3. update world and do collision / motion / AI 4. render new screen 5. send data out
3 and 4 are the processor intensive ones, and they're highly related. In fact, one or the other is likely to monopolize the CPU in a game. From anecdotal evidence I've seen, it seems the only real second thread most games use is the sound thread, which is needed mostly for scheduling and managing the sounds to play. Sound is usually not very intensive on processor cycles, especially with modern sound cards carrying most of the work.
I challenge someone not at work to take a screenshot of the Windows System manager with the threads open tab, as a simple verification here. I suspect that ut2k4 and doom3 and bf2 will all have relatively low thread counts; I don't bother with this stuff often but it would also be neat if there's a tool out there to show how the cpu time is divided across threads.
As you alluded to, this has often been attributed to Lincoln, although we don't have quite as definitive proof as the above video. I've also seen it attributed as an ancient Chinese proverb, but either way, I'm pretty sure it predates both Bush presidencies.
I'm pretty sure section 7 is designed to handle this. You can make the case that it doesn't do a good enough job, though. I'm guessing the problem is that it doesn't explicitly say a patent owner participating in the GPL must also grant a liscence.
There are a few cards out there that have slots for that purpose, and lots of server boards have tons of RAM slots, to which you could potentially run the OS off a RAM drive cache, but it's generally smarter to let the OS handle the balancing act than explicitly specify what to do. Some admins will naturally differ on this, but that's their job, to do stuff the computer can't.
In particular, the GPL (as opposed to the LGPL) only allows non-commercial software to use your copyrighted/patented work, so you can still make a good business licensing it to commercial users.
I'm not sure where you came up with that idea, but as I understand it, if you GPL something, original or derivative, you must include a liscense to use any patents it may contain. That doesn't mean you can't sell the software for commercial gain, but it usually is very difficult to make a profit under GPL conditions. Basically, the GPL allows commercial software, and it's market forces that dissuade such things.
Perhaps what is needed is simply a workforce diverse enough to include people who don't generally play games. Ironically, this has been a basic requirement for years. To make games, you must love them. Why else would you work for below industry standard wages, at above average hours? Hell, the hours itself are family unfriendly, and that alone induces a sort of imbalance on the workforce behind the design of games.
Hell, a ton of public places offer wifi for free already in KC. Most every library, coffeeshops and university offers it. Crown Center, Union Station, Hell, I think a couple of bars have public access for some reason; perhaps a public service as an apology for the shitty clientele. Not to mention that KC has some of the worst sprawl I've seen, and plenty of buildings and hills to interfere with line of sight.
But you've struck up a pretty good point. A lot of businesses won't or can't adopt wifi because it simply isn't secure. Hell, the local wifi businesses can't even guarentee their hardware is safe.
If you're having trouble seeing the city offer it up, imagine Sprint proposing it on a whim and you can't help but imagine the local officials jumping to serve their biggest employer.
Yea, that's probably what I was thinking of. Silly me. Ironically, I don't use their Linux CrossOver stuff (or other offerings) because its just simpler and cheaper to dual boot Linux. But I don't really think software companies are about to think "Sure, its reasonable to think that mac users are willing to spend extra money just to use our software."
There's tons of poor students in theatrics who can act. The trouble is getting that one male actor who doesn't exude that strange homosexual voice pattern stereotyped on TV, or is able to at least overcome it. Because pretty much every video game is about Mr The Man kicking ass, taking names and chewing bubblegum, or Madam Buxum doing those exact same things while managing a pair breasts that seem to move of their own free will.
Right now, you're quite right, Mac versus Windows questions is apple versus orange (pardon the pun). But once you can install Windows on your Mac, then it becomes possible to directly compare the two in a sensible manner. We can find out whether OSX really penalizes Apache for it's threadpool or whatever system call intensive server app you had in mind. We can also make a comparison between OSX running on a particular x86 versus PPC chipset, and perhaps say "these two differently clocked chips are comparable in most tasks, although Intel has a marginal lead in floating point operations." Its possible that we won't be able to find an intersection between old and new Mac performance, although if the prior marketing was to be believed, it should be very difficult to make an Intel chip that outperforms a G5.
Nobody really cares about UNIX guts; UNIX is a succession quick hacks. I use linux on my desktop, but it's not because I have a deepfound appreciation for the syscall interface it provides; I like it because it's cheap, it works and I have less problems with spyware.
It turns out that a handful of graphic artists need the processing power every once in a while so they can apply Photoshop filters and undo them faster, or to some goofy video editing with After Effects. Realistically they just need a faster CPU and RAM speeds, but don't make the mistake of pointing out that PCI-X is somehow inferior to the newer PCIe in some fashion, it offends their sense of importance (and really, what do corporate artist have left, if not self-importance?).
Yea, unfortunately, Best Buy is only paying their Geeks like 20 dollars an hour, tops. You could start your own tech support business, but GeekSquad can charge that much because they have a captive audience with Best Buy.
So I guess the place to sign up is Wall Street, under the sign BBY.
I was aware of the distinctions between PCI-X and PCIe. However, PCIe is not "just for graphics cards." If your ignorant inflammatory ass had spent any time researching today's technologies, you'd know by now that PCIe is slated to be the successor to PCI, and that the 2 gigabyte transfer rate of PCI-X 2.0 is about equivelent to an eight lane PCIe bus. PCIe currently goes up to 16 lanes, for a total of 4 gigabytes per second throughput. I fully expect to see RAID cards in the near future that use PCIe, and consumer video cards have already jumped on that bandwagon. That doesn't change the fact that your hardware isn't currently PCIe, naturally.
I brought up multicore simply because it was a similar technology based on the need for more CPU cycles, in parallel. Hell, for the most part, multicore is just a cost saving version of multi processing.
As for RAM, I suppose you're correct that video editing stations love the RAM. Maybe I'm just crazy, but with Dual Channel technology, you just plain can't get 3 slots anymore. Granted, that's mostly an AMD technology, and Intel chips are still far superiour for floating point operations like video compression and ray tracing. But as you correctly pointed out, most apps can't even use stuff above the four gig limit.
My simple point was that a lot of the technology in workstations is headed for the above average desktop. PCIe addresses the same problems as PCI-X and does a better job of it. RAM is facing the 64-bit problem, and in a year, the price of a 2 gig stick of ram should come down signficantly. Now maybe your professional needs will increase to compensate, but I still think that workstations and desktops are going to look more and more alike.
Finally, pull the stick out of your ass you condescending prick.
Most Americans didn't care about the Microsoft antitrust suit, and a significant number of them are shareholders either direct or through mutual funds. The American government has proven itself incapable of regulating its most influential subjects. And Microsoft's direct destruction of the browser market has had a severe impact on internet security. Not that competition can stop things like users handing out passwords for candy, but it's certainly a healthy step forward.
Of course, that was five years ago. Things have changed: * PCI-E is something like 10 times faster than PCI-X. If you're looking for PCI-X support, well that's gonna cost the same premium as every other legacy. I think most graphic artists (and that's who the PowerMac is really targeted to) can get away without it. * Memory is getting so cheap that paying extra for eight memory slots is more expensive than the savings of multiple smaller sticks of RAM, unless you were really looking for something to support like 12 gigabytes, which I would classify as server, not workstation. * Multicore processing isn't multiprocessing, but I don't expect them to start shipping Xeon workstations either. But I expect that the first x86 powerbooks will be multicore.
Workstations cost more, but the difference between a workstation and a simple desktop computer is quickly becoming marginalized.
I guess they liscenced their OS too cheaply. Apple doesn't seem able to extract itself from the hardware business. I realize and agree that there'd be a lot of initial costs getting OSX to drive the myriad of 3rd party hardware in the world, but it'd practically be a liscence to mint money once complete. Unfortunately, Apple is mostly afraid of losing their ridiculusly high markup hardware sales model to pursue this course of action, and Alienware would love to pursue them on it (although the gamer stigma would be difficult to balance).
This could signal the end of Apple's 'hardware company' identity. In the past Apple has functioned as a hardware designer and systems integrator. What I mean is they have been using a lot of engineering talent designing their motherboards and choosing the chips that go into it. PC OEMs no longer participate in that, with the gigantic number of factories pretty much doing that for them. Its possible they could simply follow suit and drop or move a bunch of their engineers off of the Mac, and torward the next Apple-life-saving iPod device.
What I don't get is the people who say Apple's real source of revenue is. The fact is that hardware is a low margin game. 10 percent is fantastic. On the other hand, the software industry is ridiciulusly high margin; the industry average is like 20 percent. Now, software might not be AAPL's best source of revenue, but it could become it's profit center in short order. If millions of emachines users suddenly ran out and bought a copy of OSXi, it's possible that Apple's revenue might fall (but with millions of sales, that's unlikely), but their profits are almost gurarenteed to take off. Even if they're just above the bottom quartile in margins, they'd still be doing better than where they are now. What funds OS development is expected profits. If Steve Jobs really thought that OSX was vital to its success, it wouldn't matter if their hardware was bringing in money or not, they'd issue bonds, and pay them off with future profit.
This process will take some time; expect about two years before they even consider opening OSX up in private business meetings. Several things will happen during this time frame. Managment will realize they won't be needing all that engineering anymore; 3rd party peripheral manufacturers will start making drivers available; OEMs will start calling up suggesting a deal on OSX. Then they'll have to begin the process of hiring new software guys and start pressuring manufacturers to release OSX drivers.
Is it reasonable though, for the Attorney General to be the one in charge of the sites to be blocked? I suspect this will be one of the core arguments of the ALCU, that the government cannot play a role in this without limiting free speech.
Slashdot Mirror
swap?
Realistically, it doesn't matter what the Bush administration wants to see done in Iran. Establishing lasting Iraqi democracy will take years, and the volunteer force we have now is being stretched thin by our Iraqi presence. The fundamental strategy of war is to go in with more forces than your opponent. And the more so, the better-- less causualties are expected when you've got ten times as many troops than one extra troop. This isn't Risk.
So we went into iraq on a high casulty mission. Fortunately, this turned out to be a non-issue; the military forces Congress was told Saddam was massing in preperation was non-existant. However, the defense department lowballed the estimated need for troops by about 60 percent. Shinseki was asking for large numbers, to reduce casulties and occupy the region. It's beginning (if not already) to appear that he was at least more correct, as attacks continue to increase, and foreign fighters begin falling into the region.
So maybe we're seeing some casulties, but it's only a small number of the hundred thousand or so deployed, right? Well, those casulties are significant in a volunteer army. Enlisted soldiers are great, but enlistment itself is violently tied to public opinion for current events and leadership, even though service commitments are longer than any elected position lasts. In a vicious cycle, that negative impact has negative reprocussions-- stop loss measures enacted to keep enlistment high. This is basically a form of drafting people highly skilled in military needs (ie former soldiers). As this becomes public news, it further erodes the desire to enlist amongst the general population.
At this point, invading Iran would require acts of severe delusion by Congressmembers, or a draft (assuming those can be distinct possibilities). The draft is the missing element to a Vietnam re-enactment in 2005. I don't suspect anyone in Congress is ready to face that sort of pressure.
The first time I added the threads tab, I was shocked that so many programs actually included threading, given how little time we spent on threading in my classes, and how much the proffessors indicated that threading was a difficult problem to solve.
I'm not sure where the extra threads come from in normal applications; perhaps there's some threading involved with the basic UI. For example, it appears that spider.exe creates a new thread to animate a new deal of cards.
At any rate, the parent is correct; games are multithreaded, I even admit this in my parent post. But they're not exactly optimizable for threads, which was my point, and which the grand-grand parent explicitly admitted. But if all the work happens in a single thread, then multiprocessing can't ever really help.
You're already somewhat ahead of the game. However, generally speaking, you've got one, maybe two intensive threads.
And your threading for a FPS is going to be highly interdependent. The process is generally:
1. get new packet
2. get input
3. update world and do collision / motion / AI
4. render new screen
5. send data out
3 and 4 are the processor intensive ones, and they're highly related. In fact, one or the other is likely to monopolize the CPU in a game. From anecdotal evidence I've seen, it seems the only real second thread most games use is the sound thread, which is needed mostly for scheduling and managing the sounds to play. Sound is usually not very intensive on processor cycles, especially with modern sound cards carrying most of the work.
I challenge someone not at work to take a screenshot of the Windows System manager with the threads open tab, as a simple verification here. I suspect that ut2k4 and doom3 and bf2 will all have relatively low thread counts; I don't bother with this stuff often but it would also be neat if there's a tool out there to show how the cpu time is divided across threads.
He did coin a similar phrase though.
"There's an old saying in Tennessee -- I know it's in Texas, probably in Tennessee -- that says, fool me once, shame on -- shame on you. Fool me -- you can't get fooled again"
As you alluded to, this has often been attributed to Lincoln, although we don't have quite as definitive proof as the above video. I've also seen it attributed as an ancient Chinese proverb, but either way, I'm pretty sure it predates both Bush presidencies.
I'm pretty sure section 7 is designed to handle this. You can make the case that it doesn't do a good enough job, though. I'm guessing the problem is that it doesn't explicitly say a patent owner participating in the GPL must also grant a liscence.
There are a few cards out there that have slots for that purpose, and lots of server boards have tons of RAM slots, to which you could potentially run the OS off a RAM drive cache, but it's generally smarter to let the OS handle the balancing act than explicitly specify what to do. Some admins will naturally differ on this, but that's their job, to do stuff the computer can't.
In particular, the GPL (as opposed to the LGPL) only allows non-commercial software to use your copyrighted/patented work, so you can still make a good business licensing it to commercial users.
I'm not sure where you came up with that idea, but as I understand it, if you GPL something, original or derivative, you must include a liscense to use any patents it may contain. That doesn't mean you can't sell the software for commercial gain, but it usually is very difficult to make a profit under GPL conditions. Basically, the GPL allows commercial software, and it's market forces that dissuade such things.
I wonder if the Guild makes him pay himself royalties?
Another favorite sign is the "DANGER! resistance in excess of 10 thousand ohms!"
Perhaps what is needed is simply a workforce diverse enough to include people who don't generally play games. Ironically, this has been a basic requirement for years. To make games, you must love them. Why else would you work for below industry standard wages, at above average hours? Hell, the hours itself are family unfriendly, and that alone induces a sort of imbalance on the workforce behind the design of games.
Hell, a ton of public places offer wifi for free already in KC. Most every library, coffeeshops and university offers it. Crown Center, Union Station, Hell, I think a couple of bars have public access for some reason; perhaps a public service as an apology for the shitty clientele. Not to mention that KC has some of the worst sprawl I've seen, and plenty of buildings and hills to interfere with line of sight.
But you've struck up a pretty good point. A lot of businesses won't or can't adopt wifi because it simply isn't secure. Hell, the local wifi businesses can't even guarentee their hardware is safe.
If you're having trouble seeing the city offer it up, imagine Sprint proposing it on a whim and you can't help but imagine the local officials jumping to serve their biggest employer.
Yea, that's probably what I was thinking of. Silly me. Ironically, I don't use their Linux CrossOver stuff (or other offerings) because its just simpler and cheaper to dual boot Linux. But I don't really think software companies are about to think "Sure, its reasonable to think that mac users are willing to spend extra money just to use our software."
How does the existance of another IDE stifle people from porting Windows apps to OSX? If anything, it should encourage more OSX software than less...
There's tons of poor students in theatrics who can act. The trouble is getting that one male actor who doesn't exude that strange homosexual voice pattern stereotyped on TV, or is able to at least overcome it. Because pretty much every video game is about Mr The Man kicking ass, taking names and chewing bubblegum, or Madam Buxum doing those exact same things while managing a pair breasts that seem to move of their own free will.
Right now, you're quite right, Mac versus Windows questions is apple versus orange (pardon the pun). But once you can install Windows on your Mac, then it becomes possible to directly compare the two in a sensible manner. We can find out whether OSX really penalizes Apache for it's threadpool or whatever system call intensive server app you had in mind. We can also make a comparison between OSX running on a particular x86 versus PPC chipset, and perhaps say "these two differently clocked chips are comparable in most tasks, although Intel has a marginal lead in floating point operations." Its possible that we won't be able to find an intersection between old and new Mac performance, although if the prior marketing was to be believed, it should be very difficult to make an Intel chip that outperforms a G5.
Nobody really cares about UNIX guts; UNIX is a succession quick hacks. I use linux on my desktop, but it's not because I have a deepfound appreciation for the syscall interface it provides; I like it because it's cheap, it works and I have less problems with spyware.
It turns out that a handful of graphic artists need the processing power every once in a while so they can apply Photoshop filters and undo them faster, or to some goofy video editing with After Effects. Realistically they just need a faster CPU and RAM speeds, but don't make the mistake of pointing out that PCI-X is somehow inferior to the newer PCIe in some fashion, it offends their sense of importance (and really, what do corporate artist have left, if not self-importance?).
Yea, unfortunately, Best Buy is only paying their Geeks like 20 dollars an hour, tops. You could start your own tech support business, but GeekSquad can charge that much because they have a captive audience with Best Buy.
So I guess the place to sign up is Wall Street, under the sign BBY.
I was aware of the distinctions between PCI-X and PCIe. However, PCIe is not "just for graphics cards." If your ignorant inflammatory ass had spent any time researching today's technologies, you'd know by now that PCIe is slated to be the successor to PCI, and that the 2 gigabyte transfer rate of PCI-X 2.0 is about equivelent to an eight lane PCIe bus. PCIe currently goes up to 16 lanes, for a total of 4 gigabytes per second throughput. I fully expect to see RAID cards in the near future that use PCIe, and consumer video cards have already jumped on that bandwagon. That doesn't change the fact that your hardware isn't currently PCIe, naturally.
I brought up multicore simply because it was a similar technology based on the need for more CPU cycles, in parallel. Hell, for the most part, multicore is just a cost saving version of multi processing.
As for RAM, I suppose you're correct that video editing stations love the RAM. Maybe I'm just crazy, but with Dual Channel technology, you just plain can't get 3 slots anymore. Granted, that's mostly an AMD technology, and Intel chips are still far superiour for floating point operations like video compression and ray tracing. But as you correctly pointed out, most apps can't even use stuff above the four gig limit.
My simple point was that a lot of the technology in workstations is headed for the above average desktop. PCIe addresses the same problems as PCI-X and does a better job of it. RAM is facing the 64-bit problem, and in a year, the price of a 2 gig stick of ram should come down signficantly. Now maybe your professional needs will increase to compensate, but I still think that workstations and desktops are going to look more and more alike.
Finally, pull the stick out of your ass you condescending prick.
Most Americans didn't care about the Microsoft antitrust suit, and a significant number of them are shareholders either direct or through mutual funds. The American government has proven itself incapable of regulating its most influential subjects. And Microsoft's direct destruction of the browser market has had a severe impact on internet security. Not that competition can stop things like users handing out passwords for candy, but it's certainly a healthy step forward.
Of course, that was five years ago. Things have changed:
* PCI-E is something like 10 times faster than PCI-X. If you're looking for PCI-X support, well that's gonna cost the same premium as every other legacy. I think most graphic artists (and that's who the PowerMac is really targeted to) can get away without it.
* Memory is getting so cheap that paying extra for eight memory slots is more expensive than the savings of multiple smaller sticks of RAM, unless you were really looking for something to support like 12 gigabytes, which I would classify as server, not workstation.
* Multicore processing isn't multiprocessing, but I don't expect them to start shipping Xeon workstations either. But I expect that the first x86 powerbooks will be multicore.
Workstations cost more, but the difference between a workstation and a simple desktop computer is quickly becoming marginalized.
I guess they liscenced their OS too cheaply. Apple doesn't seem able to extract itself from the hardware business. I realize and agree that there'd be a lot of initial costs getting OSX to drive the myriad of 3rd party hardware in the world, but it'd practically be a liscence to mint money once complete. Unfortunately, Apple is mostly afraid of losing their ridiculusly high markup hardware sales model to pursue this course of action, and Alienware would love to pursue them on it (although the gamer stigma would be difficult to balance).
No, that's Lou Albano, previously famous for male on male "professional" wrestling. Not quite the same, but they are both rather hairy beasts.
This could signal the end of Apple's 'hardware company' identity. In the past Apple has functioned as a hardware designer and systems integrator. What I mean is they have been using a lot of engineering talent designing their motherboards and choosing the chips that go into it. PC OEMs no longer participate in that, with the gigantic number of factories pretty much doing that for them. Its possible they could simply follow suit and drop or move a bunch of their engineers off of the Mac, and torward the next Apple-life-saving iPod device.
What I don't get is the people who say Apple's real source of revenue is. The fact is that hardware is a low margin game. 10 percent is fantastic. On the other hand, the software industry is ridiciulusly high margin; the industry average is like 20 percent. Now, software might not be AAPL's best source of revenue, but it could become it's profit center in short order. If millions of emachines users suddenly ran out and bought a copy of OSXi, it's possible that Apple's revenue might fall (but with millions of sales, that's unlikely), but their profits are almost gurarenteed to take off. Even if they're just above the bottom quartile in margins, they'd still be doing better than where they are now. What funds OS development is expected profits. If Steve Jobs really thought that OSX was vital to its success, it wouldn't matter if their hardware was bringing in money or not, they'd issue bonds, and pay them off with future profit.
This process will take some time; expect about two years before they even consider opening OSX up in private business meetings. Several things will happen during this time frame. Managment will realize they won't be needing all that engineering anymore; 3rd party peripheral manufacturers will start making drivers available; OEMs will start calling up suggesting a deal on OSX. Then they'll have to begin the process of hiring new software guys and start pressuring manufacturers to release OSX drivers.
Is it reasonable though, for the Attorney General to be the one in charge of the sites to be blocked? I suspect this will be one of the core arguments of the ALCU, that the government cannot play a role in this without limiting free speech.