Re:No... a 64bit chip doesn't have to be 'slower'
on
AMD's 64-Bit Chip
·
· Score: 2
Case and point. If a 64-bit processor had a 64-bit ADD instruction that takes 16 cycles to complete, a similar 32-bit ADD instruction (assuming the addends and sum are within the 32-bit range, which would be the case if performed on a strictly 32-bit app) would take approximately half (8) the amount of cycles to complete.
That depends on the design of the adder. Full adders, half adders, ripple-carry adders, and other types of adder exhibit different behavior as the word length is increased.
Consider: the upper 32 bits can be added in parallel with the lower 32 bits, and a correction step added in the case where the lower 32 bits overflow into the upper 32. That won't take twice as long as a single 64-bit add, though it will require more logic.
The audio CD standard specifies CLV (constant linear velocity) reading, so the data rate is the same at all points on the disc. At higher speeds, this method requires a very powerful motor to spin up the disc to read the inner tracks. Past a certain point, the inner tracks cannot be spun fast enough without the disc deforming to maintain the CLV method, so CAV (constant angular velocity) began to be used. This requires a smaller drive motor, as it only needs enough power to maintain rotation and a burst of power for spin-up, and is more stable, since there is no straining motor introducing vibration to the disc.
The True-X drives are P-CAV (partial constant angular velocity) drives, meaning the transfer rate ramps up from the center of the CD and more quickly reaches its maximum, where it stays throughout the majority of the disc. For more, visit storagereview.
Reality master, eh? Not the reality the rest of us share.
A company which only makes 48X burners misses out on the market for 24-40X burners. A company which makes 24X, 32X, 40X and 48X burners spends an awful lot of R&D effort designing 4 separate products. A company which makes a 48X burner, bin-splits the drives that can't quite make 48X and changes a few bits in the firmware to compete in 4 separate markets spends less money on R&D, less money on manufacturing, and can compete in more markets than a manufacturer which makes separate parts for each product.
You apparently don't know it, but the reason certain Intel CPUs always overclocked extremely well was that their manufacturing process had gotten so good there weren't enough CPUs that maxed out at the lower speeds, so they sold chips capable of 20-50% higher clock speed as lower-clocked chips in order to compete in that market. Remember that making a CPU costs a few bucks - it's the R&D and fab upgrades that cost an arm and a leg and losing marketshare to AMD is incredibly costlier than selling what might have been a $300 CPU for $85 is.
Remember those square-hole-punch devices that would "magically" turn a 720K floppy into a 1.44M floppy? They worked because it's cheaper to make 1.44M-capable media and stick it in everything than to make 1.44M and 720K media and keep them separate at the factory. The only difference between the higher and lower-capacity media WAS the hole in one corner.
I could come up with more examples, but it boils down to you being wrong. Some 32X burners won't hit 40X; some 40X burners won't hit 48X, but the hardware's the same (in some cases, in some product lines, after manufacturing tolerances outpace the market, and such provisos) in most of them.
While we're being petty, you might check a dictionary to back yourself up. A nice thick one, since the definition of "have" won't support you, so the book might as well. Have, in the sense the above poster was using it, is an accessory or helper verb.
Re:Now begins the hardest part...
on
Ogg Vorbis 1.0
·
· Score: 2
Actually, it's MPEG II Layer 3
Yup. MPEG2, just like I said. And.mpg and.mpeg and.mpeg2 are extensions I have found on audio-only files, so don't go saying they're video-only. As regards video.ogg files in winamp, did you actually read the press release? OR what you wrote? Ogg VORBIS is at 1.0. The Vorbis codec is for audio, so it naturally wouldn't know what to do with video (though I'll bet it'd play the audio track from an Ogg video stream), much like winamp doesn't know what to do with an mpeg2 video. If you're worried about it, you're welcome to name all your vorbis files with.vorbis (assuming you're using a reasonably modern OS/filesystem) so you don't confuse them with anything else.
Who are you to say that having one extension leading you to one entity to provide software and support is more confusing than having dozens of extensions (how many native MSOffice file formats have their own extension. All of 'em.) for the same product making things look completely different. Who would guess that.xls and.doc belong to the same product? IF nothing else, I search google for.ogg and get vorbis.com and xiph.org as the top two results. If I search for.xls, I get two pages in the top 200 from Microsoft, neither one telling me what the hell xls is. Searching for.doc gets me nothing from Microsoft in the top 200 results.
So tell me again, which is more useful, helpful, understandable, benefitial to the user?
Re:Now begins the hardest part...
on
Ogg Vorbis 1.0
·
· Score: 2
OT: Ogg is such a crappy name for a format, anyway. OGG stands for "hi, I'm a geek, I'm going to name what I create after fantasy characters."
You prefer.MP3, which stands for "hi, I'm an idiot; I don't know it's an MPEG2 file" then?
We've got.mpg,.mpeg,.mp2,.mp3, and probably several other file extensions for THE SAME STANDARD and you can't think of anything better to bitch about than the fact that ALL OGG FILES (audio, video, what have you) will have the SAME EXTENSION? Sheesh.
You could on the other hand say, "2.5 times faster than the nVidia card."
UM. No. How can you spend most of your post saying that 251% is 151% faster, and then go and say that 2.5x is 2.5x faster. It's 1.5x *faster* or 2.5x *as fast* just like it's 151% *faster* or 251% *as fast*. You shut your brain off a sentence too early.
...this technology seems to be behind times in speed when it finally gets released for PCs a year or two from now.
By 2005, hard drives will probably be in the vicinity of 100M/sec STR (double today's value) so the gigabit high-end quoted in the article will be quite acceptable. Where these discs will really have problems though is in access time. Modern CD and DVD drives have access times of about 85-125ms, 8-30 times slower than modern hard drives. Fortunately, discs are used primarily for STR-bound tasks such as playing audio or video, backing up data, and the like. What is important, I think, is the expense of making drives and even more so of making the discs themselves. Then there's the issue of need. CDs are starting to show their age. DVDs haven't matured yet to take their place, but are begining to feel the pinch as well. Only backup applications would need a terabyte of space and until there is a mass-market need for the technology, it will be expensive and relegated to the market currently occupied by tape drives today.
Remember, the hard drive is probably the bottleneck in almost every PC and server, particularly with huge databases. I would really like to see hard drives get faster and faster instead of bigger and bigger.
I disagree. Granted, faster is better, but it really isn't a problem. Hard drives are a mass storage device, and should not be used for a computer's working data set. This is why 32-bit addressing is such a problem - 4G isn't much data at all for a large DB. I don't know much about databases, but it seems to me many DB operations are limited by access time, not STR. And access time is limited by the moving parts inside hard drives - you can only flick a read head back and forth so fast before things start to break (and melt too). On the other hand, data storage needs increase year by year and for 99% of them, the speed of access isn't very important.
Personally, I'd rather see RAM get cheaper (and faster) than hard disks get faster. 50M/sec and 100-400 IO/sec is more than I need from a single drive. Besides, both those figures can be increased dramatically with RAID if needed. OTOH, modern CPUs are all starved for data in nearly all operations. I suspect in the future we will see more 3-level caches with the third level consisting of a fairly large amount (128M-1GB) of very fast DDR DRAM and main memory consisting of many RAM banks such as are found in modern servers. Apple's got a L3 DDR cache that delivers ~2-3G/sec, but many main memory systems can deliver that bandwidth. GFX cards on the other hand are in the 10-20G/sec range and would, I suspect, dramatically benefit high-end CPUs. See, for example, HP and IBM's MCM technology with large shared caches. Another approach may be to design an ISA with very many registers replacing the common L1 cache for efficiency, leaving only L2 and L3 caches buffering memory accesses.
"As a policy, Microsoft rarely speaks out against partners. Even when bugs in Mac OS X hampered the release of Office v. X, MacBU took the heat for product delays rather than blaming Apple."
Well, there's a brilliant piece of spindoctoring! "We've screwed up so much in the past that nobody would believe us if we blamed someone else for something that didn't work" suddenly becomes "We're such a noble company we'll take the flak to protect our allies." Masterful.
"entire music collection in your pocket"
on
Making the iPod
·
· Score: 2
Any idea when Apple's going to release a bigger iPod? Obviously Tosh will have to make a bigger drive first, but a bigger drive doesn't automatically mean a bigger iPod.
If Apple wants to live up to their advertizing, they'll need to do a damn sight better than 10G. I'm not saying there's anything wrong with the current iPod, but my music collection would fill several iPods (yeah, the "big" ones), which then would fail to fit in my pocket.
Of course it's not my decision to make. Not remotely. By extension, it's not the government's place to prosecute attempted suicides / assisted suicide, but that's another debate.
I'm afraid many of the responders to my original post think I'm a despicable human being. Perhaps I am, but not by virtue of the ideas in that post - I was trying to start a discussion.
To say that it's "okay" to trade one life for many lives assumes that you KNOW exactly what that life might have the potential to become or affect. In reality we don't know the potential or what/who that person might change that might be significant.
Exactly! So any problems you see with my hypothetical situations apply as well to yours. Sure, my hypothetical last-minute deaths could be 5 minutes too soon to save millions. But then my 5 or 20 or a thousand hypothetical lives could be absolute treasures too.
My point is that precisely because you DON'T know how any life or death will turn out, you have to work with what you do know. Trading one random peasant for 100 other random peasants increases the chance of one of your hypothetical breakthroughs by orders of magnitude.
So you say it's easy to trade one life for 50 as long as it's not the life of someone you know. And I say you're damn right it is! Not only would I trade one stranger's life to make 100 other people's lives better, I'd trade 100 strangers lives for the life of someone I care about. And I don't feel the least bit bad about it. The people I love are few and far between but to me, they're worth an airplane of passengers crashing into a mountain. Each and every one of them. I suppose it sounds bad but the truth is I care very little about people I don't know and have no connection to. 1000 people die in an earthquake somewhere, and I don't feel the slightest twinge.
But if I could turn one chance at saving the world into 50 and all I had to do were push a button, I'd press its candy red surface. Twice, if they'd let me. And while we're on the subject of hypothetical situations and the value of human life, we can philosophize about what a universal cure for cancer is worth. A life? 1000 lives? A million? While I'm being uncharitable and callus, I'll bet there are a million people whose lives will ammount to just north of nothing at all and they aren't very hard to find. Leukemia, anyone? Going once, going twice...
What moral principle of yours led to that mini-rant? That all life is sacred? Abandon that a moment, and follow me.
You don't have to put a dollar amount on a human life in order for the answer to the question to be "yes, where do I sign up?" You merely have to define an equivalence between lives. To you, are all human lives equal in value? If so, consider how many lives you could save with the $million you get by killing one innocent (though that's an entire other discussion) person.
How about a murkier situation? Suppose you can never be completely sure that you have saved someone's life. How many lives do you have to make better before it balances out one death? If you take ten people on the path to death by drug overdose of one form or another and you help them set their lives straight, are you even? If your million buys a child a medical procedure he or she needs to survive, have you erased your moral debt? After all, the peasant you killed probably had less life left than the life you granted the child. What if you set up a trust that provided scholarships (or funded cancer research, or made yearly donations to your favorite charities, or what have you)?
Moving in another direction, how would you respond if we provided some background to the initial decision? Suppose you knew the peasant in question would die tomorrow if you did nothing. Is one day of one life worth any of the above? Suppose the deaf peasant were about to be hit by a train just coming around the bend. What then? Would you sacrifice two minutes of a person's life in return for adding 20 years to someone else's life? Would you do it in exchange for a cure for HIV? Do you consider the good you could do with a million dollars morally different from a cure for HIV?
As for the windows issue, it would be cool if they could replace the walls and ceiling with polarizeable high impact plastic of some kind. Make the entire thing into a skylight! It wouldn't give you much view of the ground unless you were near one of the edges, but it would still help everyone feel less cramped i think.
Being able to sit without my knees pressed into the seat in front of me would make me feel less cramped. Being able to walk down the aisle without dodging televisions hung from the ceiling would make me feel less cramped. Not being surrounded by people who insist on bringing aboard luggage which is clearly in violation of the size restrictions on carry-on baggage would make me feel less cramped. A bathroom I can stand up straight in would make me feel less cramped. A sink big big enough to fit my hands under the faucet would make me feel less cramped. Seats with some padding so I can sink into them rather than constantly being on edge would make me feel less cramped.
I really don't think the windows have anything to do with it.
How is this not a monopoly? How about "How *IS* this a monopoly?"
Micron is the 2nd-largest DRAM manufacturer. Thus, they don't monopolize the market. They are physically incapable of producing that much DRAM, even if they had bought Nanya's guts once the banks finish with them. If Micron lowered prices, they would gain a bit of market share but would lose money. If they raise prices, module makers will ignore them. They have no control over the DRAM market. By definition, a monopolist does. I don't understand what part of that is confusing to you.
How 'bout the fact that the memory manufacturers had to sell DRAMs for *LESS* than it cost to make them for months last year? Read all about it on Cnet ( http://news.com.com/2100-1001-271208.html) - DRAM prices fell by over 90% in less than a year. They're not getting rich, they're losing their shirts. Granted, Micron's poised to be one (and not number 1) of about 3 big memory manufacturers left, but everyone playing the memory game is racing as fast as they possibly can to try to get ahead - and everyone's staying neck-and-neck. The manufacturers are making DRAMs as cheaply as possible using the best technology available and they're just ekeing out a profit.
If being one of three huge companies engaged in cut-throat competition constitutes a "monopoly" then I'd love to see a Microsoft monopoly.
Some facts: 1. Rambus was part of the JEDEC. 2. JEDEC rules require disclosure of patents. 3. Rambus declosed no patents. 4. (while part of JEDEC) Rambus suggested certain methods employed in SDRAM and DDR (that they had patents on). 5. Now that SDR/DDR is well-entrenched, Rambus is suing for royalties.
The obvious (to me) solution is to take a closer look at numbers 2 and 3. Rambus was required to declare their patents. They did not do so. Thus, they have no patents on the technology incorporated into JEDEC standards. Don't mess around with antitrust laws, just invalidate the patents in question and let them try to make a living on the technology they've got legitimate claims to.
I've got one of their original models (very limited memory, no spellcheck / IR / curvy blue plastic) that's still going strong. In almost 10 years of owning it, I think I've changed the batteries 3 times, once when I left it on for a week. I've replaced the keyboard (it just got too gummy and dusty for my taste) and the power switch (they sent me a replacement for free and I soldered it in) but it's been tossed, dropped, scratched, and neglected for 10 years and other than that switch has only cosmetic damage to show for it.
It's ugly as sin and fairly limited, but it does what it does with nothing to crash, with nonvolatile memory, and with nary a complaint. An upgraded product is definately appreciated, but somehow I don't see this new model providing the same trouble-free experience as the old standby.
It's not that the material gets so bumpy it's impossible to go further - you just have to make read heads that can read the data from higher up - the two problems aren't separate. I can't find the link right now, but IBM's press material concerning glass platters includes a micrograph of the surface of a glass platter versus an aluminum platter showing the glass platter to be 60% flatter so the read heads can fly closer to the media without fear of head crashes.
Number three *is* poorly worded; I'll try to restate it better.
Third, nanotubes aren't 10x as strong as steel. Steel is 10x as strong as steel - it just takes 10x as much of it. Nanotubes might be 10x as strong as an equivalent weight of steel, or an equivalent volume thereof, but that's not what the article says. Or maybe they meant an equivalent number of atoms, though if all you have is 1 atom, it's not steel...
From a company recently committed to supporting X86-64, this seems an odd omission. Clearly anyone buying AMD kit to avoid paying for licenses to Windows will get their asses sued before they can say "EULA" but there are plenty of cases and a strong precident for the legality of taking advantage of mistakes, errors, or omissions in contracts. If the license agreement provides licenses for all computers and charges based on the number of "Pentiums, Power Macs, iMacs or better", the decision will come down to interpretation. What does the word better refer to? Is it "(Pentiums), (Power Macs), (iMacs or better)" or is it "(Pentiums, Power Macs, iMacs) or better"? Recently, the Pentiums have become better (faster) than AMD's best chip, but I hardly think that defense will stand (though I'd be tickled pink if Gateway switched to all AMD boxen because they save $100/machine on software).
Slashdot Mirror
Case and point. If a 64-bit processor had a 64-bit ADD instruction that takes 16 cycles to complete, a similar 32-bit ADD instruction (assuming the addends and sum are within the 32-bit range, which would be the case if performed on a strictly 32-bit app) would take approximately half (8) the amount of cycles to complete.
That depends on the design of the adder. Full adders, half adders, ripple-carry adders, and other types of adder exhibit different behavior as the word length is increased.
Consider: the upper 32 bits can be added in parallel with the lower 32 bits, and a correction step added in the case where the lower 32 bits overflow into the upper 32. That won't take twice as long as a single 64-bit add, though it will require more logic.
The audio CD standard specifies CLV (constant linear velocity) reading, so the data rate is the same at all points on the disc. At higher speeds, this method requires a very powerful motor to spin up the disc to read the inner tracks. Past a certain point, the inner tracks cannot be spun fast enough without the disc deforming to maintain the CLV method, so CAV (constant angular velocity) began to be used. This requires a smaller drive motor, as it only needs enough power to maintain rotation and a burst of power for spin-up, and is more stable, since there is no straining motor introducing vibration to the disc.
The True-X drives are P-CAV (partial constant angular velocity) drives, meaning the transfer rate ramps up from the center of the CD and more quickly reaches its maximum, where it stays throughout the majority of the disc. For more, visit storagereview.
Reality master, eh? Not the reality the rest of us share.
A company which only makes 48X burners misses out on the market for 24-40X burners. A company which makes 24X, 32X, 40X and 48X burners spends an awful lot of R&D effort designing 4 separate products. A company which makes a 48X burner, bin-splits the drives that can't quite make 48X and changes a few bits in the firmware to compete in 4 separate markets spends less money on R&D, less money on manufacturing, and can compete in more markets than a manufacturer which makes separate parts for each product.
You apparently don't know it, but the reason certain Intel CPUs always overclocked extremely well was that their manufacturing process had gotten so good there weren't enough CPUs that maxed out at the lower speeds, so they sold chips capable of 20-50% higher clock speed as lower-clocked chips in order to compete in that market. Remember that making a CPU costs a few bucks - it's the R&D and fab upgrades that cost an arm and a leg and losing marketshare to AMD is incredibly costlier than selling what might have been a $300 CPU for $85 is.
Remember those square-hole-punch devices that would "magically" turn a 720K floppy into a 1.44M floppy? They worked because it's cheaper to make 1.44M-capable media and stick it in everything than to make 1.44M and 720K media and keep them separate at the factory. The only difference between the higher and lower-capacity media WAS the hole in one corner.
I could come up with more examples, but it boils down to you being wrong. Some 32X burners won't hit 40X; some 40X burners won't hit 48X, but the hardware's the same (in some cases, in some product lines, after manufacturing tolerances outpace the market, and such provisos) in most of them.
While we're being petty, you might check a dictionary to back yourself up. A nice thick one, since the definition of "have" won't support you, so the book might as well. Have, in the sense the above poster was using it, is an accessory or helper verb.
Actually, it's MPEG II Layer 3
.mpg and .mpeg and .mpeg2 are extensions I have found on audio-only files, so don't go saying they're video-only. As regards video .ogg files in winamp, did you actually read the press release? OR what you wrote? Ogg VORBIS is at 1.0. The Vorbis codec is for audio, so it naturally wouldn't know what to do with video (though I'll bet it'd play the audio track from an Ogg video stream), much like winamp doesn't know what to do with an mpeg2 video. If you're worried about it, you're welcome to name all your vorbis files with .vorbis (assuming you're using a reasonably modern OS/filesystem) so you don't confuse them with anything else.
.xls and .doc belong to the same product? IF nothing else, I search google for .ogg and get vorbis.com and xiph.org as the top two results. If I search for .xls, I get two pages in the top 200 from Microsoft, neither one telling me what the hell xls is. Searching for .doc gets me nothing from Microsoft in the top 200 results.
Yup. MPEG2, just like I said. And
Who are you to say that having one extension leading you to one entity to provide software and support is more confusing than having dozens of extensions (how many native MSOffice file formats have their own extension. All of 'em.) for the same product making things look completely different. Who would guess that
So tell me again, which is more useful, helpful, understandable, benefitial to the user?
OT: Ogg is such a crappy name for a format, anyway. OGG stands for "hi, I'm a geek, I'm going to name what I create after fantasy characters."
.MP3, which stands for "hi, I'm an idiot; I don't know it's an MPEG2 file" then?
.mpg, .mpeg, .mp2, .mp3, and probably several other file extensions for THE SAME STANDARD and you can't think of anything better to bitch about than the fact that ALL OGG FILES (audio, video, what have you) will have the SAME EXTENSION? Sheesh.
You prefer
We've got
I'll eat my hat.....
You could on the other hand say, "2.5 times faster than the nVidia card."
UM. No. How can you spend most of your post saying that 251% is 151% faster, and then go and say that 2.5x is 2.5x faster. It's 1.5x *faster* or 2.5x *as fast* just like it's 151% *faster* or 251% *as fast*. You shut your brain off a sentence too early.
Xenon is an element. Xeon is an expensive CPU. I see "Intel Xenon" too many times at work. Please not on Slashdot too.
...this technology seems to be behind times in speed when it finally gets released for PCs a year or two from now.
By 2005, hard drives will probably be in the vicinity of 100M/sec STR (double today's value) so the gigabit high-end quoted in the article will be quite acceptable. Where these discs will really have problems though is in access time. Modern CD and DVD drives have access times of about 85-125ms, 8-30 times slower than modern hard drives. Fortunately, discs are used primarily for STR-bound tasks such as playing audio or video, backing up data, and the like. What is important, I think, is the expense of making drives and even more so of making the discs themselves. Then there's the issue of need. CDs are starting to show their age. DVDs haven't matured yet to take their place, but are begining to feel the pinch as well. Only backup applications would need a terabyte of space and until there is a mass-market need for the technology, it will be expensive and relegated to the market currently occupied by tape drives today.
Remember, the hard drive is probably the bottleneck in almost every PC and server, particularly with huge databases. I would really like to see hard drives get faster and faster instead of bigger and bigger.
I disagree. Granted, faster is better, but it really isn't a problem. Hard drives are a mass storage device, and should not be used for a computer's working data set. This is why 32-bit addressing is such a problem - 4G isn't much data at all for a large DB. I don't know much about databases, but it seems to me many DB operations are limited by access time, not STR. And access time is limited by the moving parts inside hard drives - you can only flick a read head back and forth so fast before things start to break (and melt too). On the other hand, data storage needs increase year by year and for 99% of them, the speed of access isn't very important.
Personally, I'd rather see RAM get cheaper (and faster) than hard disks get faster. 50M/sec and 100-400 IO/sec is more than I need from a single drive. Besides, both those figures can be increased dramatically with RAID if needed. OTOH, modern CPUs are all starved for data in nearly all operations. I suspect in the future we will see more 3-level caches with the third level consisting of a fairly large amount (128M-1GB) of very fast DDR DRAM and main memory consisting of many RAM banks such as are found in modern servers. Apple's got a L3 DDR cache that delivers ~2-3G/sec, but many main memory systems can deliver that bandwidth. GFX cards on the other hand are in the 10-20G/sec range and would, I suspect, dramatically benefit high-end CPUs. See, for example, HP and IBM's MCM technology with large shared caches. Another approach may be to design an ISA with very many registers replacing the common L1 cache for efficiency, leaving only L2 and L3 caches buffering memory accesses.
"As a policy, Microsoft rarely speaks out against partners. Even when bugs in Mac OS X hampered the release of Office v. X, MacBU took the heat for product delays rather than blaming Apple."
Well, there's a brilliant piece of spindoctoring! "We've screwed up so much in the past that nobody would believe us if we blamed someone else for something that didn't work" suddenly becomes "We're such a noble company we'll take the flak to protect our allies." Masterful.
Any idea when Apple's going to release a bigger iPod? Obviously Tosh will have to make a bigger drive first, but a bigger drive doesn't automatically mean a bigger iPod.
If Apple wants to live up to their advertizing, they'll need to do a damn sight better than 10G. I'm not saying there's anything wrong with the current iPod, but my music collection would fill several iPods (yeah, the "big" ones), which then would fail to fit in my pocket.
Of course it's not my decision to make. Not remotely. By extension, it's not the government's place to prosecute attempted suicides / assisted suicide, but that's another debate.
I'm afraid many of the responders to my original post think I'm a despicable human being. Perhaps I am, but not by virtue of the ideas in that post - I was trying to start a discussion.
No they couldn't:)
To say that it's "okay" to trade one life for many lives assumes that you KNOW exactly what that life might have the potential to become or affect. In reality we don't know the potential or what/who that person might change that might be significant.
Exactly! So any problems you see with my hypothetical situations apply as well to yours. Sure, my hypothetical last-minute deaths could be 5 minutes too soon to save millions. But then my 5 or 20 or a thousand hypothetical lives could be absolute treasures too.
My point is that precisely because you DON'T know how any life or death will turn out, you have to work with what you do know. Trading one random peasant for 100 other random peasants increases the chance of one of your hypothetical breakthroughs by orders of magnitude.
So you say it's easy to trade one life for 50 as long as it's not the life of someone you know. And I say you're damn right it is! Not only would I trade one stranger's life to make 100 other people's lives better, I'd trade 100 strangers lives for the life of someone I care about. And I don't feel the least bit bad about it. The people I love are few and far between but to me, they're worth an airplane of passengers crashing into a mountain. Each and every one of them. I suppose it sounds bad but the truth is I care very little about people I don't know and have no connection to. 1000 people die in an earthquake somewhere, and I don't feel the slightest twinge.
But if I could turn one chance at saving the world into 50 and all I had to do were push a button, I'd press its candy red surface. Twice, if they'd let me. And while we're on the subject of hypothetical situations and the value of human life, we can philosophize about what a universal cure for cancer is worth. A life? 1000 lives? A million? While I'm being uncharitable and callus, I'll bet there are a million people whose lives will ammount to just north of nothing at all and they aren't very hard to find. Leukemia, anyone? Going once, going twice...
What moral principle of yours led to that mini-rant? That all life is sacred? Abandon that a moment, and follow me.
You don't have to put a dollar amount on a human life in order for the answer to the question to be "yes, where do I sign up?" You merely have to define an equivalence between lives. To you, are all human lives equal in value? If so, consider how many lives you could save with the $million you get by killing one innocent (though that's an entire other discussion) person.
How about a murkier situation? Suppose you can never be completely sure that you have saved someone's life. How many lives do you have to make better before it balances out one death? If you take ten people on the path to death by drug overdose of one form or another and you help them set their lives straight, are you even? If your million buys a child a medical procedure he or she needs to survive, have you erased your moral debt? After all, the peasant you killed probably had less life left than the life you granted the child. What if you set up a trust that provided scholarships (or funded cancer research, or made yearly donations to your favorite charities, or what have you)?
Moving in another direction, how would you respond if we provided some background to the initial decision? Suppose you knew the peasant in question would die tomorrow if you did nothing. Is one day of one life worth any of the above? Suppose the deaf peasant were about to be hit by a train just coming around the bend. What then? Would you sacrifice two minutes of a person's life in return for adding 20 years to someone else's life? Would you do it in exchange for a cure for HIV? Do you consider the good you could do with a million dollars morally different from a cure for HIV?
As for the windows issue, it would be cool if they could replace the walls and ceiling with polarizeable high impact plastic of some kind. Make the entire thing into a skylight! It wouldn't give you much view of the ground unless you were near one of the edges, but it would still help everyone feel less cramped i think.
Being able to sit without my knees pressed into the seat in front of me would make me feel less cramped. Being able to walk down the aisle without dodging televisions hung from the ceiling would make me feel less cramped. Not being surrounded by people who insist on bringing aboard luggage which is clearly in violation of the size restrictions on carry-on baggage would make me feel less cramped. A bathroom I can stand up straight in would make me feel less cramped. A sink big big enough to fit my hands under the faucet would make me feel less cramped. Seats with some padding so I can sink into them rather than constantly being on edge would make me feel less cramped.
I really don't think the windows have anything to do with it.
How is this not a monopoly?
How about "How *IS* this a monopoly?"
Micron is the 2nd-largest DRAM manufacturer. Thus, they don't monopolize the market. They are physically incapable of producing that much DRAM, even if they had bought Nanya's guts once the banks finish with them. If Micron lowered prices, they would gain a bit of market share but would lose money. If they raise prices, module makers will ignore them. They have no control over the DRAM market. By definition, a monopolist does. I don't understand what part of that is confusing to you.
How 'bout the fact that the memory manufacturers had to sell DRAMs for *LESS* than it cost to make them for months last year? Read all about it on Cnet ( http://news.com.com/2100-1001-271208.html) - DRAM prices fell by over 90% in less than a year. They're not getting rich, they're losing their shirts. Granted, Micron's poised to be one (and not number 1) of about 3 big memory manufacturers left, but everyone playing the memory game is racing as fast as they possibly can to try to get ahead - and everyone's staying neck-and-neck. The manufacturers are making DRAMs as cheaply as possible using the best technology available and they're just ekeing out a profit.
If being one of three huge companies engaged in cut-throat competition constitutes a "monopoly" then I'd love to see a Microsoft monopoly.
Some facts:
1. Rambus was part of the JEDEC.
2. JEDEC rules require disclosure of patents.
3. Rambus declosed no patents.
4. (while part of JEDEC) Rambus suggested certain methods employed in SDRAM and DDR (that they had patents on).
5. Now that SDR/DDR is well-entrenched, Rambus is suing for royalties.
The obvious (to me) solution is to take a closer look at numbers 2 and 3. Rambus was required to declare their patents. They did not do so. Thus, they have no patents on the technology incorporated into JEDEC standards. Don't mess around with antitrust laws, just invalidate the patents in question and let them try to make a living on the technology they've got legitimate claims to.
I've got one of their original models (very limited memory, no spellcheck / IR / curvy blue plastic) that's still going strong. In almost 10 years of owning it, I think I've changed the batteries 3 times, once when I left it on for a week. I've replaced the keyboard (it just got too gummy and dusty for my taste) and the power switch (they sent me a replacement for free and I soldered it in) but it's been tossed, dropped, scratched, and neglected for 10 years and other than that switch has only cosmetic damage to show for it.
It's ugly as sin and fairly limited, but it does what it does with nothing to crash, with nonvolatile memory, and with nary a complaint. An upgraded product is definately appreciated, but somehow I don't see this new model providing the same trouble-free experience as the old standby.
It's not that the material gets so bumpy it's impossible to go further - you just have to make read heads that can read the data from higher up - the two problems aren't separate. I can't find the link right now, but IBM's press material concerning glass platters includes a micrograph of the surface of a glass platter versus an aluminum platter showing the glass platter to be 60% flatter so the read heads can fly closer to the media without fear of head crashes.
Number three *is* poorly worded; I'll try to restate it better.
Third, nanotubes aren't 10x as strong as steel. Steel is 10x as strong as steel - it just takes 10x as much of it. Nanotubes might be 10x as strong as an equivalent weight of steel, or an equivalent volume thereof, but that's not what the article says. Or maybe they meant an equivalent number of atoms, though if all you have is 1 atom, it's not steel...
Why, oh why can't people write?
Nah. A terabyte is 5 bytes (2^40 requires 5 bytes to index).
From a company recently committed to supporting X86-64, this seems an odd omission. Clearly anyone buying AMD kit to avoid paying for licenses to Windows will get their asses sued before they can say "EULA" but there are plenty of cases and a strong precident for the legality of taking advantage of mistakes, errors, or omissions in contracts. If the license agreement provides licenses for all computers and charges based on the number of "Pentiums, Power Macs, iMacs or better", the decision will come down to interpretation. What does the word better refer to? Is it "(Pentiums), (Power Macs), (iMacs or better)" or is it "(Pentiums, Power Macs, iMacs) or better"? Recently, the Pentiums have become better (faster) than AMD's best chip, but I hardly think that defense will stand (though I'd be tickled pink if Gateway switched to all AMD boxen because they save $100/machine on software).