Hot-swap. The ground connectors are longer than the power connectors. This grounds the drive's electronics before power is applied - prevents potential differences from destroying delicate parts.
> why not old Molex?
Friction-fit Molex power connectors suck. Just ask anyone who has used one more than 5 times.
The new SATA power and data connectors allow the drive to be hot-swapped with a minimum of extras. The drives can be slid into protective cases or hot-swapped bare - a vast improvement over the bulky boxes required for current parallel IDE drives to achieve even warm-swapping.
It doesn't matter a whit what the warranty period is if your only concern is data integrity. The hard drive *will* fail. It *will* fail at the least opportune time. It *will* take your data with it. What'cha gonna do about it? RAID of course. And if you've got the choice between 3 drives at $500 each or 7 drives at $140 each + RAID controller (RAID10 with hot spare), and you choose the $500 drives, you're fired.
Putting your data in only one place is *never* the right way if it's at all important to you. Your data goes in one place so you can work on it. It goes in another place so the hardware can fail (redundancy to protect against a dead HD). It goes in a third place so the people can fail (Oops! I deleted a file three weeks ago and just noticed!). It goes in a fourth place so reality can fail (lightning, flood, etc.). And if you're paranoid, it goes in a fifth place in case something unforseen happens.
Yes, it will cost you at least $500 for as much storage space as you could get for that $140, but that doesn't mean buying a $500 hard drive in the first place is the right answer.
What bugs you about OO? I've noticed some quirks here and there (It does not play well with multimonitor setups, for instance), but it fixes some of my biggest gripes with MS Office. I've been very impressed with the MS Office and Powerpoint import filters as well (certainly better than Abiword, which is what drove me to buy a student-edition copy of Office for $25. It's not even installed anymore.). Are you just a heavier user than I and have noticed more rough edges?
I don't use WMP under Windows - why would I want it cluttering up my linux box? I suspect the sort of people that actually like MS office would jump at the chance to run WMP under linux. I don't trust my audio and video files to a company which can't make a decent web browser, email client, word processor, or presentation program, and actively attacks those who can.
I know I'm in the minority here, but I've used everything from word95 to word2000, and it wasn't until OpenOffice came along that I abandoned WordPerfect 5.1 for DOS.
Since for once I don't have mod points to give away, I thought I'd comment instead.
As someone with a 10.5" hand spread (about the 98 or 99th percentile, for those who're counting), I definately echo your "what the hell?!?" upon seeing that dinky keyboard floating in an ocean of wasted space. I'll grant that the backlit keyboard looks awesome, but for Apple to spend their engineering talent making a pretty keyboard over a functional one is both in character and unforgivable. I'm not asking for a separate number keypad and pgup/pgdn/home/end/ins/del block, but a full-sized set of function keys, arrow keys, and the like is an absolute must. There was enough room for both on the old powerbooks, and there's even more space on the 17" behemoth.
True. That energy is captured by the electric field however. The particle's origination point is known, and its energy (kinetic) is known from the reaction which produces it. This lets you calibrate the electric field so that it saps the majority of the particle's kinetic energy, leaving it just enough to barely touch the walls of the reaction chamber and suck off some electrons. You're not actually wasting the kinetic energy.
Re:Ummm... what's the deal with the special power
on
Serial ATA, Here and Now
·
· Score: 5, Informative
New because the old friction-fit molex power connectors suck.
Actually, it's for hot-swapability. The old molex power connectors would make your drive virtually glued in, and you'd have to jiggle it to get the contacts to fit. The new power connectors are designed for hot-swap operation. They're smaller, easier to slide in and out, and have longer ground wires which ensures the drive is grounded *before* any power is delivered. The same long-short wiring is used in the data cables, where the 3 grounding pins connect before the 4 data pins (two pairs using differential signalling) connect.
I'm not sure how you figure an adapter is less reliable. Have you *ever* had a molex power connector come apart on you unexpectedly? I count myself lucky if I can get them apart on purpose!
This early in the migration, there may be issues here and there, but when SATA becomes the standard, there will be connectors for it right on the PSU's cabling, and motherboards will support tons of SATA channels straight to the northbridge rather than ganged onto the PCI bus, and maybe hotswap drives will start to be the norm. Alright, not that last bit, but the first two should happen pretty quickly since SATA is cheaper for the manufacturer, as well as better for the consumer.
Matter-antimatter reactions produce gamma rays and other high-energy radiation. In order to harness this energy, you need to convert it into electricity, which requires actually absorbing the radiation. But since gamma rays laugh at lead or gold shielding and blast right through, there's a wee problem.
In contrast, the device mentioned in the article produces alpha particles (when configured appropriately, using Boron fuel). Alpha particles, if they touch metals, suck off 2 electrons to become helium atoms. This produces a net charge, and voila - electricity. The use of alpha particles in this way (such as from radioactive decay of certain isotopes) is well-tested. Since the majority (perhaps 95%) of the energy produced would be in the form of alpha particles, this type of reactor has the potential to be extremely efficient.
Regrettably, I don't have the background to determine whether it's all a crock or not. It sounds plausible, but all the best ones do. I'll believe it when it's powering my computer, but I'd donate a dollar to see if it could be done.
Technically, "Moon" is the proper name for Earth's major satellite. In common usage though, a moon is any natural satellite of a planet. I have been unable to find specific requirements as to size of satellite, eccentricity of orbit, stability of orbit, etc. which would cause random bits of debris, such as those making up Saturn's rings, to be classified as moons or not.
My personal criteria would require uniqueness and stability. That is, tiny rocks in a ring are not unique, whereas a single (or two, or 40) orbiting body deserves special recognition. Also, the orbit must be stable. That is, an asteroid captured by a planet's gravity in an orbit that will smash it into something in 200 years is not a moon. It is a pet.
Are you kidding? You can see your coffee mug from space! And that's not even lit up; it has to rely on reflected light. This guy's radiating a few dozen amps of power and you're worried if it's visible from space? I'll bet we start hearing complaints from Alpha Centauri in about 8 years.
Re:Slashdot can teach us many things
on
1.5 TB DVD by 2010
·
· Score: 4, Funny
At least you don't see "that's equivalent to a stack of paper stretching from the earth to the sun 12 times" in magazines anymore.
Two cents, one for the government, one for the recipient. I get about 7,500-10,000 spams a year; at 1 cent per, that would buy me a new CPU every year. Which I could definately appreciate, as this $50-three-years-ago-Duron is starting to feel a bit threadbare.
Depends on what you set. Mozilla has a long list of javascript permissions you can turn on or off. I believe the "nix unrequested popups" box ignores all new windows unless they're created by an onClick event. It sounds like these ads are of the "onMouseOver = 'go somewhere stupid'" type. The Mozilla option to disallow javascript changing the page you are at would fix that too.
All modern CPUs have highly accurate branch predictors. Just as caches need to catch the vast majority of memory accesses (About 98% I believe is a common number for the L1+L2 caches) for the CPU to work even remotely close to its theoretical maximum, branch predictors need to avoid pipeline flushes whenever possible. Consider the P4 with 20-odd stages, more in the FP pipeline. There could be dozens of instructions in flight by the time the branch test gets completed and says it took the wrong branch - The 20-stage pipeline needs to be emptied of invalid instructions, all of which get thrown away. A new memory location needs to be loaded, and execution needs to resume at that point. Extremely inefficient, so the occurence of this sort of thing has to be kept to an absolute minimum. Since the P4 suffers so horribly in the case of mispredicted branches, it tends to do poorly in branch-heavy code such as chess benchmarks. In order to keep such wasteful operations from happening, the CPU keeps track of thousands of previous branches in an attempt to guess correctly which way a new branch will go. There is a data structure in the CPU called the BHT, Branch History Table, that holds all this information, some times with many bits of info per branch.
I won't take the time to dig up references to see if any CPU architecture currently does all of these tricks, but consider all the things that can be done to minimize the impact of branches: When a branch occurs, fetch instructions from *both* possible branch locations. Begin executing both sets of instructions in parallel, keeping the CPU's back-end busy. Flag these instructions with a "left branch" or "right branch" tag. When the branch test completes, toss out the wrong instructions and keep the good ones. Both branches will execute more slowly than a correctly-guessed branch that executes only one, but in the case of a mispredict, there is no pipeline flush, and no delay waiting for the PC to update and new instructions to flow in. Also, it's hard on the caches and RAM subsystem, since two sets of instructions rather than 1 need to be fetched. Build a better predictor. By analyzing the type of branch and surrounding code, the branch predictor can get eerily accurate. Way better than 90%. A K6-2 could get 90% accuracy with no sweat, and that's a pretty old chip. Prefetch. Grab the branch test and run it way ahead of the branch itself (when possible). If the outcome of the branch can be determined before the branch is reached (using instruction reordering trickery), there is effectively no branch at all. This can be "stacked" with other techniques as well. I'm sure you can come up with several more. It's an interesting problem to think about, with most techniques having a good mix of benefits and drawbacks.
I could have sworn it was illegal (or at least against some shrinkwrap EULA mumbo-jumbo) to play a DVD over any sort of wireless link. It came up during Microsoft's massively ill-conceived tablet PC thing, I believe.
They had some older burners and bought new as new technology came out? Wow, you could almost make that into a business model!
Re:Linux more popular than Microsoft
on
Web Zeitgeist
·
· Score: 2
Nah. Microsoft's not scared of linux at #4 or ftp at #5. They're pissed that realplayer beat them out, and they're gleeful at having beaten java. And look at xbox at #7. For a console that's practically rotting on the shelves, that's pretty durn good. Looks like maybe there *is* a light at the end of that $3 billion tunnel.
I'm sorry, but that's not +5 informative, it's -1 misinformed.
First, some terms.
CLV = Constant Linear Velocity. The original audio CD standard specified a constant datarate, which meant that the disc spun faster at the center tracks and slower at the outside. This was fine, since CDs mostly played linearly through, and even if you switched tracks, there was time for the drive motor to adjust the speed.
CAV = Constant Angular Velocity. The disc spins at the same RPMs no matter what part is being read. This requires a much less powerful drive motor and allows for random access.
There are others too, such as P-CAV (partial CAV) and Z-CLV (zone CLV), but the main distinction is between CLV and CAV.
Suppose a CD burner were to operate in CLV mode like the AC proposes. The inner tracks, being 1/6 as long as the outer tracks, would have to spin very fast to achieve the desired datarate. But, CDs cannot spin much past 10K RPM or they will self-destruct (and some do anyway). So, if the inner tracks are spinning at 10K, the outer tracks must spin at 10K/6 = about 1500RPM in order to maintain the constant linear velocity. But the drive motor and media can keep up with 10K RPM speeds, so why limit them to 1500RPM?
If CD burners were CAV on the other hand, the drive would maintain a constant 10K RPM and the burner would fire the laser for a slightly different durration each time it writes a pit so the pits end up being the same length on the final CD. But that's hard to do over a continously variable linear speed (RPMs * track length = linear speed). So, the burners use Z-CLV. The drive motor spins up to 10K and writing starts; it slows down slightly to compensate for the tracks lengthening, so the burner can write each pit the same length. When the zone ends, the burner again ramps up to 10K RPM and the process continues at a higher datarate.
The AC seems to think that 24X means the disc is spinning slower than 30X, and he's being cheated on the inner tracks which start writing slowly. That's just not true. The disc spins just as fast on the inner zone as it does on the outer zone (and maybe faster, as there's less turbulence and disc wiggle there!), but the tracks are so short that not much data gets written per revolution.
It is true that substantially more bits are dedicated to ECC under ISO9660 than under the CD-AUDIO standard. However, this does not imply that audio CDs are more succeptible to error. Just the opposite, in fact. CDs were never intended to be a random-access format, so when ISO9660 came along and wanted random access, considerable effort went into making it work, such as the extra ECC and positioning information in subchannels. Moreover, data requires that each bit be reproduced accurately. Audio has no such requirement. So, when an audio CD player cannot read a bit, it performs correction by interpolating from the previous and next samples on the disc (or by some other, more accurate, method). This cannot be done with data, which is another reason why so much ECC is required for data CDs. Have you noticed that you can run an audio CD over in your car, buff it out a bit, and play through just fine? Ever tried that with the CD you write your backups to?
I hope I've been informative. And moderators? Please don't moderate what you don't understand!
Slashdot Mirror
> why new?
Hot-swap. The ground connectors are longer than the power connectors. This grounds the drive's electronics before power is applied - prevents potential differences from destroying delicate parts.
> why not old Molex?
Friction-fit Molex power connectors suck. Just ask anyone who has used one more than 5 times.
The new SATA power and data connectors allow the drive to be hot-swapped with a minimum of extras. The drives can be slid into protective cases or hot-swapped bare - a vast improvement over the bulky boxes required for current parallel IDE drives to achieve even warm-swapping.
Wait, the day after you started, or the day after you finished?
Simple maths? Yup.
It doesn't matter a whit what the warranty period is if your only concern is data integrity. The hard drive *will* fail. It *will* fail at the least opportune time. It *will* take your data with it. What'cha gonna do about it? RAID of course. And if you've got the choice between 3 drives at $500 each or 7 drives at $140 each + RAID controller (RAID10 with hot spare), and you choose the $500 drives, you're fired.
Putting your data in only one place is *never* the right way if it's at all important to you. Your data goes in one place so you can work on it. It goes in another place so the hardware can fail (redundancy to protect against a dead HD). It goes in a third place so the people can fail (Oops! I deleted a file three weeks ago and just noticed!). It goes in a fourth place so reality can fail (lightning, flood, etc.). And if you're paranoid, it goes in a fifth place in case something unforseen happens.
Yes, it will cost you at least $500 for as much storage space as you could get for that $140, but that doesn't mean buying a $500 hard drive in the first place is the right answer.
Did a flamebait AC post actually use a homonym properly? Will wonders never cease!
What bugs you about OO? I've noticed some quirks here and there (It does not play well with multimonitor setups, for instance), but it fixes some of my biggest gripes with MS Office. I've been very impressed with the MS Office and Powerpoint import filters as well (certainly better than Abiword, which is what drove me to buy a student-edition copy of Office for $25. It's not even installed anymore.). Are you just a heavier user than I and have noticed more rough edges?
I don't use WMP under Windows - why would I want it cluttering up my linux box? I suspect the sort of people that actually like MS office would jump at the chance to run WMP under linux. I don't trust my audio and video files to a company which can't make a decent web browser, email client, word processor, or presentation program, and actively attacks those who can.
I know I'm in the minority here, but I've used everything from word95 to word2000, and it wasn't until OpenOffice came along that I abandoned WordPerfect 5.1 for DOS.
And when the feature gets added, send an email, a postcard, a box of chocolates, The Almighty Buck, or a picture of your cat. Say thank you.
Since for once I don't have mod points to give away, I thought I'd comment instead.
As someone with a 10.5" hand spread (about the 98 or 99th percentile, for those who're counting), I definately echo your "what the hell?!?" upon seeing that dinky keyboard floating in an ocean of wasted space. I'll grant that the backlit keyboard looks awesome, but for Apple to spend their engineering talent making a pretty keyboard over a functional one is both in character and unforgivable. I'm not asking for a separate number keypad and pgup/pgdn/home/end/ins/del block, but a full-sized set of function keys, arrow keys, and the like is an absolute must. There was enough room for both on the old powerbooks, and there's even more space on the 17" behemoth.
Whoops. Alsee's response better addresses the issue. I should have read more before posting.
True. That energy is captured by the electric field however. The particle's origination point is known, and its energy (kinetic) is known from the reaction which produces it. This lets you calibrate the electric field so that it saps the majority of the particle's kinetic energy, leaving it just enough to barely touch the walls of the reaction chamber and suck off some electrons. You're not actually wasting the kinetic energy.
New because the old friction-fit molex power connectors suck.
Actually, it's for hot-swapability. The old molex power connectors would make your drive virtually glued in, and you'd have to jiggle it to get the contacts to fit. The new power connectors are designed for hot-swap operation. They're smaller, easier to slide in and out, and have longer ground wires which ensures the drive is grounded *before* any power is delivered. The same long-short wiring is used in the data cables, where the 3 grounding pins connect before the 4 data pins (two pairs using differential signalling) connect.
I'm not sure how you figure an adapter is less reliable. Have you *ever* had a molex power connector come apart on you unexpectedly? I count myself lucky if I can get them apart on purpose!
This early in the migration, there may be issues here and there, but when SATA becomes the standard, there will be connectors for it right on the PSU's cabling, and motherboards will support tons of SATA channels straight to the northbridge rather than ganged onto the PCI bus, and maybe hotswap drives will start to be the norm. Alright, not that last bit, but the first two should happen pretty quickly since SATA is cheaper for the manufacturer, as well as better for the consumer.
Matter-antimatter reactions produce gamma rays and other high-energy radiation. In order to harness this energy, you need to convert it into electricity, which requires actually absorbing the radiation. But since gamma rays laugh at lead or gold shielding and blast right through, there's a wee problem.
In contrast, the device mentioned in the article produces alpha particles (when configured appropriately, using Boron fuel). Alpha particles, if they touch metals, suck off 2 electrons to become helium atoms. This produces a net charge, and voila - electricity. The use of alpha particles in this way (such as from radioactive decay of certain isotopes) is well-tested. Since the majority (perhaps 95%) of the energy produced would be in the form of alpha particles, this type of reactor has the potential to be extremely efficient.
Regrettably, I don't have the background to determine whether it's all a crock or not. It sounds plausible, but all the best ones do. I'll believe it when it's powering my computer, but I'd donate a dollar to see if it could be done.
Technically, "Moon" is the proper name for Earth's major satellite. In common usage though, a moon is any natural satellite of a planet. I have been unable to find specific requirements as to size of satellite, eccentricity of orbit, stability of orbit, etc. which would cause random bits of debris, such as those making up Saturn's rings, to be classified as moons or not.
My personal criteria would require uniqueness and stability. That is, tiny rocks in a ring are not unique, whereas a single (or two, or 40) orbiting body deserves special recognition. Also, the orbit must be stable. That is, an asteroid captured by a planet's gravity in an orbit that will smash it into something in 200 years is not a moon. It is a pet.
Are you kidding? You can see your coffee mug from space! And that's not even lit up; it has to rely on reflected light. This guy's radiating a few dozen amps of power and you're worried if it's visible from space? I'll bet we start hearing complaints from Alpha Centauri in about 8 years.
At least you don't see "that's equivalent to a stack of paper stretching from the earth to the sun 12 times" in magazines anymore.
is *my* hero!
Two cents, one for the government, one for the recipient. I get about 7,500-10,000 spams a year; at 1 cent per, that would buy me a new CPU every year. Which I could definately appreciate, as this $50-three-years-ago-Duron is starting to feel a bit threadbare.
Depends on what you set. Mozilla has a long list of javascript permissions you can turn on or off. I believe the "nix unrequested popups" box ignores all new windows unless they're created by an onClick event. It sounds like these ads are of the "onMouseOver = 'go somewhere stupid'" type. The Mozilla option to disallow javascript changing the page you are at would fix that too.
Or you could just turn javascript off completely.
All modern CPUs have highly accurate branch predictors. Just as caches need to catch the vast majority of memory accesses (About 98% I believe is a common number for the L1+L2 caches) for the CPU to work even remotely close to its theoretical maximum, branch predictors need to avoid pipeline flushes whenever possible. Consider the P4 with 20-odd stages, more in the FP pipeline. There could be dozens of instructions in flight by the time the branch test gets completed and says it took the wrong branch - The 20-stage pipeline needs to be emptied of invalid instructions, all of which get thrown away. A new memory location needs to be loaded, and execution needs to resume at that point. Extremely inefficient, so the occurence of this sort of thing has to be kept to an absolute minimum. Since the P4 suffers so horribly in the case of mispredicted branches, it tends to do poorly in branch-heavy code such as chess benchmarks. In order to keep such wasteful operations from happening, the CPU keeps track of thousands of previous branches in an attempt to guess correctly which way a new branch will go. There is a data structure in the CPU called the BHT, Branch History Table, that holds all this information, some times with many bits of info per branch.
I won't take the time to dig up references to see if any CPU architecture currently does all of these tricks, but consider all the things that can be done to minimize the impact of branches:
When a branch occurs, fetch instructions from *both* possible branch locations. Begin executing both sets of instructions in parallel, keeping the CPU's back-end busy. Flag these instructions with a "left branch" or "right branch" tag. When the branch test completes, toss out the wrong instructions and keep the good ones. Both branches will execute more slowly than a correctly-guessed branch that executes only one, but in the case of a mispredict, there is no pipeline flush, and no delay waiting for the PC to update and new instructions to flow in. Also, it's hard on the caches and RAM subsystem, since two sets of instructions rather than 1 need to be fetched.
Build a better predictor. By analyzing the type of branch and surrounding code, the branch predictor can get eerily accurate. Way better than 90%. A K6-2 could get 90% accuracy with no sweat, and that's a pretty old chip.
Prefetch. Grab the branch test and run it way ahead of the branch itself (when possible). If the outcome of the branch can be determined before the branch is reached (using instruction reordering trickery), there is effectively no branch at all. This can be "stacked" with other techniques as well.
I'm sure you can come up with several more. It's an interesting problem to think about, with most techniques having a good mix of benefits and drawbacks.
I could have sworn it was illegal (or at least against some shrinkwrap EULA mumbo-jumbo) to play a DVD over any sort of wireless link. It came up during Microsoft's massively ill-conceived tablet PC thing, I believe.
> "See Dick. See Jane. See Dick run."
You did read the review, right? "See Dick. See Jane. See Jane fuck Dick. See Dick die three days later. Poor Dick."
Ok. Let me see if I've got this straight. I fill out that form and the RIAA will give me some money to buy a DVD with?
They had some older burners and bought new as new technology came out? Wow, you could almost make that into a business model!
Nah. Microsoft's not scared of linux at #4 or ftp at #5. They're pissed that realplayer beat them out, and they're gleeful at having beaten java. And look at xbox at #7. For a console that's practically rotting on the shelves, that's pretty durn good. Looks like maybe there *is* a light at the end of that $3 billion tunnel.
I'm sorry, but that's not +5 informative, it's -1 misinformed.
First, some terms.
CLV = Constant Linear Velocity. The original audio CD standard specified a constant datarate, which meant that the disc spun faster at the center tracks and slower at the outside. This was fine, since CDs mostly played linearly through, and even if you switched tracks, there was time for the drive motor to adjust the speed.
CAV = Constant Angular Velocity. The disc spins at the same RPMs no matter what part is being read. This requires a much less powerful drive motor and allows for random access.
There are others too, such as P-CAV (partial CAV) and Z-CLV (zone CLV), but the main distinction is between CLV and CAV.
Suppose a CD burner were to operate in CLV mode like the AC proposes. The inner tracks, being 1/6 as long as the outer tracks, would have to spin very fast to achieve the desired datarate. But, CDs cannot spin much past 10K RPM or they will self-destruct (and some do anyway). So, if the inner tracks are spinning at 10K, the outer tracks must spin at 10K/6 = about 1500RPM in order to maintain the constant linear velocity. But the drive motor and media can keep up with 10K RPM speeds, so why limit them to 1500RPM?
If CD burners were CAV on the other hand, the drive would maintain a constant 10K RPM and the burner would fire the laser for a slightly different durration each time it writes a pit so the pits end up being the same length on the final CD. But that's hard to do over a continously variable linear speed (RPMs * track length = linear speed). So, the burners use Z-CLV. The drive motor spins up to 10K and writing starts; it slows down slightly to compensate for the tracks lengthening, so the burner can write each pit the same length. When the zone ends, the burner again ramps up to 10K RPM and the process continues at a higher datarate.
The AC seems to think that 24X means the disc is spinning slower than 30X, and he's being cheated on the inner tracks which start writing slowly. That's just not true. The disc spins just as fast on the inner zone as it does on the outer zone (and maybe faster, as there's less turbulence and disc wiggle there!), but the tracks are so short that not much data gets written per revolution.
It is true that substantially more bits are dedicated to ECC under ISO9660 than under the CD-AUDIO standard. However, this does not imply that audio CDs are more succeptible to error. Just the opposite, in fact. CDs were never intended to be a random-access format, so when ISO9660 came along and wanted random access, considerable effort went into making it work, such as the extra ECC and positioning information in subchannels. Moreover, data requires that each bit be reproduced accurately. Audio has no such requirement. So, when an audio CD player cannot read a bit, it performs correction by interpolating from the previous and next samples on the disc (or by some other, more accurate, method). This cannot be done with data, which is another reason why so much ECC is required for data CDs. Have you noticed that you can run an audio CD over in your car, buff it out a bit, and play through just fine? Ever tried that with the CD you write your backups to?
I hope I've been informative. And moderators? Please don't moderate what you don't understand!