Do you have any idea what you are talking about? "Hardware abstraction"? All kernels do hardware abstraction, what does that have to do with multithreading?
Regardless of your architecture or kernel, if an application only starts one thread, it is normally impossible to run that thread on two processors at once. You can run two different applications like that on two processors at once, but not one on both, which is usually what the task is.
And the software doesn't need to be recompiled in any case (except of course if you want to make it multithreaded so it can take advantage of it). It's the same code, the OS takes care of loading it into different processors and making everything happy.
I mean, I use XP, and turning off the Big Brother features is as easy as getting a corporate edition (ok, not very easy for those who want to actually buy it) and turning off a bunch of services, setting a few policies, and uninstalling Messenger after install. Oh, and configure Media Player and IE, or just don't use them.
After that, it's pretty much the same as win2k in terms of spyware. In other areas, though, it has some notable improvements.
I expected something more involving from the film
on
A Beautiful Mind
·
· Score: 4, Interesting
After two weeks of reading praise in the reviews, I finally went to see the movie. I must say I didn't like it, possibly because it hits too close to home.
Watching Nash's life suddenly reveal itself as an empty shell, a madman's delusion, was too painful. It creeped me out so much that I lost interest in the rest of the film and the recovery to normal life that he made. I guess I became afraid of what it would be like to lose control of your mind in this manner, a very disturbing perspective.
Needless to say, beyond the amount forced upon me by the movie, I could not sympathize with the character much because of the pride and prejudice and contempt and even, I would say, malice in his competitiveness (while he had it), that he touts.
To summarize, I felt sorry for him, but even more repulsed by him, and thus by the movie.
As for his portrayal as a mathematician, it had both parts that I liked and those that I didn't. There wasn't much specifics to it though, predictably.
Windows XP supports PPPoE out of the box with no need to install any other software. For Windows 9x/NT/2K you can download RASPPPoE, a 95K PPPoE driver that is installed as a network interface by running an.inf file.
CPU at the bottom: But it has to be next to the power supply. Cooler CPUs (2 or 3 years ago) could even do without their own fan, faring on just the fan on the power supply that would suck air through their heatsink. A more radical option would be to put the cpu and agp at the bottom (and like I say in my other post in this thread, on expansion cards in the backplane), put the power supply in front of them, put the drive bays above the power supply, and the rest of the cards behind the drive bays (into the backplane too). That greatly decreases size, and if done right, allows to have one huge fan in the PS that blows air over both the power supply components and the CPU/GPU cards (that would be quieter than 3 or 4 fans there are in there now).
Then to make it even more fun make the drive bays screwless (rails with latches that you push on both sides to slide it out) and the expansion cards screwless too (rails on the side that has external access and friction in the slot on the backplane), make small power connectors (5V only?) for the drives, or give them power through a beefed-up firewire or serial ATA, and of course do away with the wide IDE/SCSI type ribbon cables. Make the chassis cover removable by pushing on latches on the sides also. Have a handle on top...
interesting point in 4... I think it could drive down the size very much... let's see...
you could have slots on it for the cpu/chipset/controllers, memory, video/AGP, firewire/ata/scsi controllers, and the rest of the cards, then stick the power supply next to the cpu/agp slots so it blows the fan on them, and stick the drive bays next to the power supply against the rest of the cards, and switch to slim 5.25" drive bays already since the CDROM form factor drives are the only thing that needs them and they can be made in the slim factor.
If they do the radiators for cpu/agp correctly they could even make it work with one big fan (4 inch diameter or something) on the power supply sucking air from behind cpu/gpu and out, and still have enough cooling power to dissipate the 100 something watts they generate. It would definitely be more serviceable since you would just take off the cover and pull out one of the cards, or slide out one of the bays.
The mobo could then be 2*8=16 to 22 cm wide (the clearance between PCI cards is currently something like 2 cm, double that for CPU, and at least 4 more for gpu/memory/controllers/sound?), maybe 20 cm deep (that's how deep most PCI cards are nowadays), and up to 15 cm high... and there could be 3 slim 5.25" drive bays and 2 full 3.5" bays. All in all this could fit in a 30x25x20 cm cube...
I'm thinking of buying the Leadtek one... I was under the impression that PNY was kind of a no-name brand, and LeadTek costs only about 15 bucks more. How has it been for you?
BTW I suspect the PNY card is built on the nvidia reference design so that's what led you to think they're all the same - but the Leadtek one modifies the reference design (ti200 but not ti500).
I'm not sure if the motherboard's/controller's power circuits would be able to handle the significant spin-up current for the hard drives... what's the current rating for firewire channels? It would probably be enough to drive them while on, but not at start-up, but either way, it would rule out having something like 4 hdds attached to one controller, since that would be an overload at start-up for sure... of course, it would be a nice feature for 5400 rpm drives to have both the power connector and the firewire connector and not to require the power connector for operation, for example to have a longer spin-up time when powered only by firewire...
This approach may also be more prone to failure than full backups (although if backup space is your limit you really have no other option). An error in one archive volume may disrupt your entire restore operation, and this has happened to me before. Then of course there are archive formats that have recovery records.
Yawn. Tell me again how the AC's post was "utter BS". You seem to have a problem with only one of his points, about plenum cable, and using my herd mentality, I can see that paying more for plenum rated cable in a situation where it doesn't have practical advantage is not wise. But then you just go off spouting dirt in direction of all the people in the discussion, while none of them provoked you. You make me sick.
I don't have time to dig up more information right now, but I believe that first, An-225 or whatever plane they used to transport Buran was just a heavily modified bomber, so it wasn't designed entirely for Buran/Energia, and second, it wasn't used just to transport Buran, but also entire parts of Energia. So it wasn't as inefficient to develop and operate as you think.
This is not a flame. Honestly. I just want to see how good macos 10 is and whether it can justify buying a mac.
But what, from what you described, is there that your mac can do and a winxp/win2k running on a recent dell laptop can't?
Let's see. Automatic network configuration in multiple environments, including modem, 100baseT, 802.11, auto-config - check.
Most of your favorite command-line utils - check (Cygwin, although parts of it are bastardized, I admit). For the rest, set up X-forwarding using xwin32 or a number of other X servers and connect to your local Unix server. Many essential IDE tools like Emacs, gcc, various java tools, already run on win32.
Cuteness - check for winxp, or at least there will be. Luna looks like crap (except for the taskbar), but hopefully we'll see some extreme cuteness once the themes community gets rolling.
If we could just refrain from the religious/monopoly/bully arguments and view the systems based on their technical merits alone, how else is macosx better?
This is a rather unfounded theory, but it's possible to imagine this.
The fourth plane that crashed on sept. 11 was about 5 minutes away from the closest nuclear plant on its course to DC. It, too, had an engine break off and crash in a separate location. The government has been very secretive about that crash. There is data that a fighter was following it at the time of the crash, and based on that we can only theorize what really happened.
It is possible to think that this plane was hijacked, not responding, and a fighter was dispatched to take it down. The credibility of this theory would largely be defined by verified data about the plane's route before the crash - if it suddenly changed its course after going some ways over the water, this would be credible.
The same mode of failure as the Penn plane may indicate the same fate: a Sidewinder missile (IR-guided) will target an engine and rip it off the wing; also, a cannon (the "cleanest" way to take it out) would probably also be shot at the engine/wing.
What the AC said. In addition to 1/r^2, there's a dielectric constant that air has that dissipates the wave's energy.
(I'm not sure how big the dissipation in the air is compared to the effect of 1/r^2 though... I'm pretty sure that all the deep space probes get picked up by the DSN, not by the retransmission satellites around the Earth.)
Unrelated question: How does the positioning system for the Cassini antenna work? For other deep space probes, are there backup antennas that are not directional? How likely is it that the positioning system for the antenna fails and then there's no way for the probe to contact Earth?
Uh... I didn't know the SSMEs were pulled, rebuilt and tested... disassemble and check the blades, yes, but this seems to imply something more complete. Can you give me a link to where the process is described?
Your keystrokes are correct. However, I personally prefer to be able to write the equation the way I see it on paper. I know I can learn to translate it into RPN quickly, but what's the point? I still type faster on my 89 than I will be able to translate some 10- or 15-term formula into RPN and enter it - and on the 89, I just enter it directly the way I see it on paper (it shows up in pretty print, too, making it that much easier to see if you entered it incorrectly).
As far as I know, _every_ calculator in existence uses RPN internally (or something that is analogous in concept to RPN). 89 may have a more complex instruction stack, but what it comes down to is put one operand in one register, next operand in another, and the opcode elsewhere (form an instruction word with it), then execute the instruction word. For complex operations, it will have to execute a lot of instructions to get the answer.
Discussing how this is similar to RPN is not very productive, however I do find TI's regular notation more useful than RPN. 89 comes closer than any other calc to allowing you to enter exactly what you have on paper and get a pretty-print answer.
All in all, it's sad to see HP go, since 48/49 are apparently at least on par with the 89, however I use my 89 every day and I think it really is a powerful platform. At least the ability to write and compile C code into an executable up to 200 K in size and have it run on the 89 is hardly matched by others.
Another remark: I think both TIs and HPs are horribly overpriced. I know that they put some money into writing the software for it, but $150 for the 89 or $170 for the 48/49? Handhelds sold for half this price have several times more memory, faster processor, and higher res display! Unfortunately, HP's demise will only make the situation worse, reducing TI's competition to zero and allowing it to set the prices at will.
BTW, what can cause these "skips" to occur?
Bad riping?
Probably a frame misalignment. Like if it's ever been downloaded off of one source and then halted and resumed from another (which is a different rip) it will be misaligned. Or if you got it corrupted somehow, then it's supposed to skip the frame that doesn't decode right and interpolate.
I didn't know Nullsoft ever produced a Mac port of winamp... can you give more information?
Re:Perhaps more interesting...
on
Kursk Finally Lifted
·
· Score: 3, Informative
You've confused Granit with Shkval.
What you've linked to are the Granit anti-ship cruise missiles. Those are the vertically-launched ramjet missiles, taken to ramjet speeds by a booster, and capable (or so the military says) of flying under the radar height in formation, then one rises up, turns its radar on, and dispatches the targets to others. Those are supposed to be capable of taking out an entire aircraft carrier battle group.
What possibly sank Kursk are the Shkval torpedoes - the supercavitating supersonic torpedoes that use jet propulsion and swim in a straight line to their target.
Both are supposed to be the most advanced missile and torpedo technologies in use by Russians, both have not been fully tested.
Do you think Russia would think twice about conquering Iraq if it weren't for the US? Do you think the Russians would be any more friendly to civilian targets in Iraq? Think Chechnya.
Yes, I do. Why do you consider the Russian nation's self-conscience is less than that of the United States? There have been some terrible mistakes in Russian history, and that's what made people much more sensitive and reserved in such problems. Yes, the Russian military right now is not in its best shape, however you cannot conclude from this that Russia would behave any worse in a solitary superpower position than the US.
Chechnya? The first campaign was indeed one of those terrible mistakes. Many civilians were killed, and corruption caused suffering. After pulling out, and after Chechnya started terrorist insurrections into neighboring lands, the military got their shit together and responded effectively and rather professionally. The second campaign was a success, and Chechnya has started the process of recovery. (Yes, there were some civilians killed by the federals in the campaign, but the number afaik is close to 100, compared to many thousands of guerilla fighters.)
Slashdot Mirror
Do you have any idea what you are talking about? "Hardware abstraction"? All kernels do hardware abstraction, what does that have to do with multithreading?
Regardless of your architecture or kernel, if an application only starts one thread, it is normally impossible to run that thread on two processors at once. You can run two different applications like that on two processors at once, but not one on both, which is usually what the task is.
And the software doesn't need to be recompiled in any case (except of course if you want to make it multithreaded so it can take advantage of it). It's the same code, the OS takes care of loading it into different processors and making everything happy.
"Embedded Passport thing"?
I mean, I use XP, and turning off the Big Brother features is as easy as getting a corporate edition (ok, not very easy for those who want to actually buy it) and turning off a bunch of services, setting a few policies, and uninstalling Messenger after install. Oh, and configure Media Player and IE, or just don't use them.
After that, it's pretty much the same as win2k in terms of spyware. In other areas, though, it has some notable improvements.
After two weeks of reading praise in the reviews, I finally went to see the movie. I must say I didn't like it, possibly because it hits too close to home.
Watching Nash's life suddenly reveal itself as an empty shell, a madman's delusion, was too painful. It creeped me out so much that I lost interest in the rest of the film and the recovery to normal life that he made. I guess I became afraid of what it would be like to lose control of your mind in this manner, a very disturbing perspective.
Needless to say, beyond the amount forced upon me by the movie, I could not sympathize with the character much because of the pride and prejudice and contempt and even, I would say, malice in his competitiveness (while he had it), that he touts.
To summarize, I felt sorry for him, but even more repulsed by him, and thus by the movie.
As for his portrayal as a mathematician, it had both parts that I liked and those that I didn't. There wasn't much specifics to it though, predictably.
RASPPPoE
Use Roaring Penguin for Linux.
CPU at the bottom: But it has to be next to the power supply. Cooler CPUs (2 or 3 years ago) could even do without their own fan, faring on just the fan on the power supply that would suck air through their heatsink. A more radical option would be to put the cpu and agp at the bottom (and like I say in my other post in this thread, on expansion cards in the backplane), put the power supply in front of them, put the drive bays above the power supply, and the rest of the cards behind the drive bays (into the backplane too). That greatly decreases size, and if done right, allows to have one huge fan in the PS that blows air over both the power supply components and the CPU/GPU cards (that would be quieter than 3 or 4 fans there are in there now).
Then to make it even more fun make the drive bays screwless (rails with latches that you push on both sides to slide it out) and the expansion cards screwless too (rails on the side that has external access and friction in the slot on the backplane), make small power connectors (5V only?) for the drives, or give them power through a beefed-up firewire or serial ATA, and of course do away with the wide IDE/SCSI type ribbon cables. Make the chassis cover removable by pushing on latches on the sides also. Have a handle on top...
interesting point in 4... I think it could drive down the size very much... let's see...
you could have slots on it for the cpu/chipset/controllers, memory, video/AGP, firewire/ata/scsi controllers, and the rest of the cards, then stick the power supply next to the cpu/agp slots so it blows the fan on them, and stick the drive bays next to the power supply against the rest of the cards, and switch to slim 5.25" drive bays already since the CDROM form factor drives are the only thing that needs them and they can be made in the slim factor.
If they do the radiators for cpu/agp correctly they could even make it work with one big fan (4 inch diameter or something) on the power supply sucking air from behind cpu/gpu and out, and still have enough cooling power to dissipate the 100 something watts they generate. It would definitely be more serviceable since you would just take off the cover and pull out one of the cards, or slide out one of the bays.
The mobo could then be 2*8=16 to 22 cm wide (the clearance between PCI cards is currently something like 2 cm, double that for CPU, and at least 4 more for gpu/memory/controllers/sound?), maybe 20 cm deep (that's how deep most PCI cards are nowadays), and up to 15 cm high... and there could be 3 slim 5.25" drive bays and 2 full 3.5" bays. All in all this could fit in a 30x25x20 cm cube...
I'm thinking of buying the Leadtek one... I was under the impression that PNY was kind of a no-name brand, and LeadTek costs only about 15 bucks more. How has it been for you?
BTW I suspect the PNY card is built on the nvidia reference design so that's what led you to think they're all the same - but the Leadtek one modifies the reference design (ti200 but not ti500).
I'm not sure if the motherboard's/controller's power circuits would be able to handle the significant spin-up current for the hard drives... what's the current rating for firewire channels? It would probably be enough to drive them while on, but not at start-up, but either way, it would rule out having something like 4 hdds attached to one controller, since that would be an overload at start-up for sure... of course, it would be a nice feature for 5400 rpm drives to have both the power connector and the firewire connector and not to require the power connector for operation, for example to have a longer spin-up time when powered only by firewire...
What was the geforce3 you bought for it?
This approach may also be more prone to failure than full backups (although if backup space is your limit you really have no other option). An error in one archive volume may disrupt your entire restore operation, and this has happened to me before. Then of course there are archive formats that have recovery records.
Yawn. Tell me again how the AC's post was "utter BS". You seem to have a problem with only one of his points, about plenum cable, and using my herd mentality, I can see that paying more for plenum rated cable in a situation where it doesn't have practical advantage is not wise. But then you just go off spouting dirt in direction of all the people in the discussion, while none of them provoked you. You make me sick.
Why go to all this trouble? Isn't it possible to just kill the process and then delete the .scr file?
I don't have time to dig up more information right now, but I believe that first, An-225 or whatever plane they used to transport Buran was just a heavily modified bomber, so it wasn't designed entirely for Buran/Energia, and second, it wasn't used just to transport Buran, but also entire parts of Energia. So it wasn't as inefficient to develop and operate as you think.
This is not a flame. Honestly. I just want to see how good macos 10 is and whether it can justify buying a mac.
But what, from what you described, is there that your mac can do and a winxp/win2k running on a recent dell laptop can't?
Let's see. Automatic network configuration in multiple environments, including modem, 100baseT, 802.11, auto-config - check.
Most of your favorite command-line utils - check (Cygwin, although parts of it are bastardized, I admit). For the rest, set up X-forwarding using xwin32 or a number of other X servers and connect to your local Unix server. Many essential IDE tools like Emacs, gcc, various java tools, already run on win32.
Cuteness - check for winxp, or at least there will be. Luna looks like crap (except for the taskbar), but hopefully we'll see some extreme cuteness once the themes community gets rolling.
If we could just refrain from the religious/monopoly/bully arguments and view the systems based on their technical merits alone, how else is macosx better?
This is a rather unfounded theory, but it's possible to imagine this.
The fourth plane that crashed on sept. 11 was about 5 minutes away from the closest nuclear plant on its course to DC. It, too, had an engine break off and crash in a separate location. The government has been very secretive about that crash. There is data that a fighter was following it at the time of the crash, and based on that we can only theorize what really happened.
It is possible to think that this plane was hijacked, not responding, and a fighter was dispatched to take it down. The credibility of this theory would largely be defined by verified data about the plane's route before the crash - if it suddenly changed its course after going some ways over the water, this would be credible.
The same mode of failure as the Penn plane may indicate the same fate: a Sidewinder missile (IR-guided) will target an engine and rip it off the wing; also, a cannon (the "cleanest" way to take it out) would probably also be shot at the engine/wing.
What the AC said. In addition to 1/r^2, there's a dielectric constant that air has that dissipates the wave's energy.
(I'm not sure how big the dissipation in the air is compared to the effect of 1/r^2 though... I'm pretty sure that all the deep space probes get picked up by the DSN, not by the retransmission satellites around the Earth.)
Unrelated question: How does the positioning system for the Cassini antenna work? For other deep space probes, are there backup antennas that are not directional? How likely is it that the positioning system for the antenna fails and then there's no way for the probe to contact Earth?
Uh... I didn't know the SSMEs were pulled, rebuilt and tested... disassemble and check the blades, yes, but this seems to imply something more complete. Can you give me a link to where the process is described?
Your keystrokes are correct. However, I personally prefer to be able to write the equation the way I see it on paper. I know I can learn to translate it into RPN quickly, but what's the point? I still type faster on my 89 than I will be able to translate some 10- or 15-term formula into RPN and enter it - and on the 89, I just enter it directly the way I see it on paper (it shows up in pretty print, too, making it that much easier to see if you entered it incorrectly).
As far as I know, _every_ calculator in existence uses RPN internally (or something that is analogous in concept to RPN). 89 may have a more complex instruction stack, but what it comes down to is put one operand in one register, next operand in another, and the opcode elsewhere (form an instruction word with it), then execute the instruction word. For complex operations, it will have to execute a lot of instructions to get the answer.
Discussing how this is similar to RPN is not very productive, however I do find TI's regular notation more useful than RPN. 89 comes closer than any other calc to allowing you to enter exactly what you have on paper and get a pretty-print answer.
All in all, it's sad to see HP go, since 48/49 are apparently at least on par with the 89, however I use my 89 every day and I think it really is a powerful platform. At least the ability to write and compile C code into an executable up to 200 K in size and have it run on the 89 is hardly matched by others.
Another remark: I think both TIs and HPs are horribly overpriced. I know that they put some money into writing the software for it, but $150 for the 89 or $170 for the 48/49? Handhelds sold for half this price have several times more memory, faster processor, and higher res display! Unfortunately, HP's demise will only make the situation worse, reducing TI's competition to zero and allowing it to set the prices at will.
Probably a frame misalignment. Like if it's ever been downloaded off of one source and then halted and resumed from another (which is a different rip) it will be misaligned. Or if you got it corrupted somehow, then it's supposed to skip the frame that doesn't decode right and interpolate.
I didn't know Nullsoft ever produced a Mac port of winamp... can you give more information?
You've confused Granit with Shkval.
What you've linked to are the Granit anti-ship cruise missiles. Those are the vertically-launched ramjet missiles, taken to ramjet speeds by a booster, and capable (or so the military says) of flying under the radar height in formation, then one rises up, turns its radar on, and dispatches the targets to others. Those are supposed to be capable of taking out an entire aircraft carrier battle group.
What possibly sank Kursk are the Shkval torpedoes - the supercavitating supersonic torpedoes that use jet propulsion and swim in a straight line to their target.
Both are supposed to be the most advanced missile and torpedo technologies in use by Russians, both have not been fully tested.
I would mod this up if I had mod points... probably the most insightful reflection of the Iraq/Palestine's mistakes I've seen in this thread.
Yes, I do. Why do you consider the Russian nation's self-conscience is less than that of the United States? There have been some terrible mistakes in Russian history, and that's what made people much more sensitive and reserved in such problems. Yes, the Russian military right now is not in its best shape, however you cannot conclude from this that Russia would behave any worse in a solitary superpower position than the US.
Chechnya? The first campaign was indeed one of those terrible mistakes. Many civilians were killed, and corruption caused suffering. After pulling out, and after Chechnya started terrorist insurrections into neighboring lands, the military got their shit together and responded effectively and rather professionally. The second campaign was a success, and Chechnya has started the process of recovery. (Yes, there were some civilians killed by the federals in the campaign, but the number afaik is close to 100, compared to many thousands of guerilla fighters.)