Exactly, and as it stands right now this is already happening.
Scientific computing has gone from exhaustive and detailed simulation to exhaustive analyses of an entire parameter space with the advent of new life science branches such as Genomics, Proteomics and all the other broadly based omics-type endeavors. This means embarrassingly parallel computation at a massive scale. At my institution we keep roughly 3000 cores humming around the clock without any difficulty, largely using legacy code.
At home I have a recording studio where the bulk of the processing is happening through plugins. Again, embarrassingly parallel workload, very easy to keep a number of cores working at 100% TODAY.
That's just two examples, but I could mention many more, and so could many of you I'm sure. I fail to see why today's software cannot make productive use of multicore architectures. Unless we're talking productivity apps, which don't use much CPU anyway...
I thought the same initially. But consider a world where one of these wireless charging technologies becomes a widely used standard. Good-bye gadgets, hello universal recharging! It's all about standards, simple ones if possible. Wireless just adds another convenience factor.
Now if we could also standardize wall outlets worldwide...
That, my friend, is a dumb idea. My Skype connection to my parents overseas, for example, is based on a P2P protocol. Giving it low QoS would badly hurt it. What I'm saying with this is you can't foresee what the next grat app is going to do under the hood, and you don't want to shoot yourself in the foot at all. Keep QoS out of the hands of most, if not all people, because if you let them choke things they will. Guaranteed!
Why are you so down on the concept? Looking for what does NOT work may be just as illuminating as our more normal way of filtering through the vast amount of sequence information. I was involved in a research effort which asked the same basic question as the one discussed here, except that we didn't dare to think of nullomers as "lethal" or "toxic". Our thoughts were more along the lines of chemically or structurally instable. We did find a large number of peptides not represented in the published databases, but the sheer number and diversity even at the tetramer level made us abandon the analysis.
So, basically, it's one regexp and a database lookup. Which is fine (how else would you do it?) but all this requires is one afternoon of PhD time followed by a lot of computer crunching.
Completely correct. I've done this exact calculation/database lookup a few months ago, and grep was the tool. Due to the size of the dataset, a flexible divide-and-conquer approach was needed, however, and it took a bit more than one afternoon of Ph.D. time to get it to work reliably. In the end, though, it wasn't really rocket science.
What bugs me in TFA are all the buzzwords and the focus on high level biological mechanisms. Sure, some sequences might be lethal or may cause immune reactions, but the simpler and much more feasible explanation would be that certain sequences just aren't chemically stable, at the DNA, RNA or protein level. My chemist coworker had all sorts of theories.
But to be honest, and not to be troll, I found Mac OS X to be relatively stupified in comparison to other OS's that I've used. OS X is pretty and all, but I prefer 'functional' over 'shiny' and I like to really dig into the inner workings of the OS that I use. OS X doesn't entirely appeal to my demographic for that reason, which is why it appeals to people who just want to use a computer and have it work without having to mess with it any more than they need to.
I find this a really odd mix of statements. I've beeen using versions of Unix for possibly longer than you've been around, which I think gives me a decent perspective of what makes an OS "shiny", "functional", useful or whatever else you may want of it. Currently I'm a very happy and productive OSX user. The last time I got excited about a box under my table was in the mid 90s over a decent SGI. Since then it's been boring and/or tedious, especially during the three years I used a Linux PC as my primary desktop. I love messing in the internals of an OS, but only when I want to. At other times I have better things to do. My younger colleagues seem to feel similarly, so I fail to see where you take your claims from.
As a parent of college age kids, may I suggest you might have made a few mistakes? I'm willing to bet that half of your upfront expenses could have been deferred or lowered. It does take some planning, determination and plain old "just say no", however. Life can be really good, if you want it. It's up to you to make it happen.
I'd have to agree at least in part. The higher up a guy is on the totempole the less information they utter, and replace it with buzzwords, filler and marketing fluff. Much of this interview was like this. Still, the first page had some interest: how the two companies started to talk, customer-centered.
I find your post rather callous. While you may be right that breakthroughs don't happen without associated risks and the occasional negative or outright dangerous result, I believe we've been extremely careless during the 20th century. Your Xray example is a good one. Physicists and biologists knew fairly early on Xray radiation was ionizing, but for quite some time it didn't occur to anyone to not expose themselves or others to high doses. How hard would it be to remain a bit cautious? And maybe save a few lives and make countless other better in the process.
TFA simply advocates caution and diligent research into negative consequences of nanotech while the technology is being developed. TFA never urges abandoning anything. I agree with the author that we should keep close tabs on this stuff and watch it for long term effects.
You're correct, the iPod has these features, but they are minor items and never get in the way of accessing and playing the music, photos and video. What I meant is that Apple demonstrates that you can resist the urge to overload and clutter the interface, onscreen as well as button count and placement.
[em]Given Apple's track record, I'd say they're the most likely candidates to figure out an elegant cell phone interface, and I'm looking forward to the iPhone for that reason alone.[/em] You're right on the money here. Apple continue to show other makers how to not overload a music player. If they can pull off a sleek and functional phone, that would force the other manufacturers back to their drawing boards. Job's credo that you cannot design a product via focus groups still holds.
What people forget, and what the study with all its flaws doesn't touch on either is the mental energy cost of switching from paying attention to the actual task to reorganizing the workspace and then back to the task. I'm not a cognitive psychologist or whatever other discipline would have the tools to quantify task switching costs, but I have seen and felt many times how I and people I have worked with loose inordinate amounts of energy to task switching.
I'd love to see a study which addresses this and tries to quantify the productivity increase one can gain from minimizing task switching. I am convinced that in the context of using large displays this is bigger than Amdahl's law as you invoked, even though your argument is certainly valid.
I sort of understand the concern too, but the reality is what makes me furious. OLGA has a ton of bad quality stuff, I agree with you on that. The official music publishers, however, also put out a ton of shit, maybe even more. Very often, sheet music for popular music is so dumbed down it's not even funny. And quite often it's not in the key it was recorded, so once the poor amateur has worked out the song, she can't even play along with the CD. That's after shelling out $3 - $5 for single song sheet music or $15 - $20 for an entire album's worth. How awful is that?
In my experience (and talking about popular music from 1960 onwards only, not jazz or classical), the only official sheet music worth the asking price is the complete Beatles book. Most everything else is a waste of your time and money. I wish OLGA and all similar sites the best of luck in their battles. They possibly do less disservice to the song writers' art than the 'official' guys do.
Total cost of ownership. The MacOSX releases really add up over the years, while you only had to pay once (but much more) for XP. Of course, you also had to buy antivirus software, muck with popup blockers and get a spyware removal tool if you were using XP, so who am I kidding?
Interesting analysis, and I have to agree to an extent. I am a Unix nerd (scientific computing) and have been an SGI customer for most of the 90s. Then commodity hardware running Linux became fast enough and had enough graphics horsepower that a switch was a nobrainer, as you say.
However, I switched away from Linux PCs on the desktop, and so have many other Unix-centered scientists, and we now use Apple computers. True, most of us like for the computer to Stay Out Of The Way, but most of us still like to be smart about computers. We just got tired of having to put the same pieces together over and over again. In today's world we can focus on getting work done, whether that be heavy duty simulations, visualization, searching databases or the web, writing up results in off the shelf office software, or prepring the next seminar in standard slideware. All at the same time, and without funny glitches. However, should Apple ever misstep we would move away in a heartbeat, unlike traditional Mac users.
On the HPC side it's much the same. We still have an SGI Altix system which can't be beat in terms of scalability of our main applications, but we have 10x as many x86-64 CPUs in a dumb cluster for the simpler 'gotta get 10,000 of these calculations done' jobs. Vendor and even Linux distro don't matter at all, we buy whatever works and comes cheap enough. Very different from the old days when we benchmarked half a dozen vendors' proprietary Unix systems for several months before settling on one, and then spent several more months in 'friendly user mode'.
The Core Duo/Solo Mac Mini and MacBooks run faster if you pair RAM sticks due to the memory architecture. This is most felt on the Minis and MacBooks (non-Pro) because they use system RAM as video RAM. From the factory all these Macs come with single RAM sticks.
Great example! Now imagine this multiplied a thousandfold or more, up and down the aisles. People will stop going there, period. See what I mean?
Re:He missed the big one...
on
Inescapable Data
·
· Score: 2, Interesting
That's on top of the ability to forge the data in these databases. Have you ever tried to tell someone at a bank that the information they have about your credit is wrong? They'll believe the computer over you any day. The computer is always right.
That's our salvation eventually. The data will be so shitty and useless as to not matter much anymore. For all the talk about realtime and linked this and that, I have a sense that much of the data will be nearly immutable and thus unable to reflect the real world. Hence useless.
Quote: Honestly, I believe that this view of "the convergence" is as overly optimistic as the 1950's notion that by now we'd be travelling in flying cars, have robots cleaning our house*, and atomic power was going to make electricity too cheap to meter. In real life there's too much (friction? drag? entropy?) due to the sheer scale and complexity of legacy systems for things to happen the way the dreamers envision.
I was thinking the same. Look at TIA and other large scale programs. They seem to be mostly useless for their intended purpose. I think we'll end up with vast quantities of bad data from which interested parties will try to divine "insight". My local mom-and-pop corner store does better than that by simply using their eyes, ears and brains. Old tech, but works great on a local scale.
Agreed! We need to keep in mind, too, that with the exception of classical and some jazz, all music is recorded as mostly mono tracks, processed over and over, and finally "spatialized" with pan, delay and reverb. Whether stereo or surround, it's all fake room information, no matter how good the underlying technology is. Like you, I think the current music surround craze is totally manufactured.
I completey agree with you, and would go even further: Most of today's music doesn't even have real LEFT and RIGHT information either. Most instrument and vocal tracks start out in MONO, and the left/right information is added via judicious application of pan and reverb and maybe delay. In other words, it's all fake. This technique does give us a pleasing separation of instruments and voices, which is great, and it gives a sense of space, which is necessary. But true spatial information it is not.
Now throw in today's craze to mix as surround sound. Same general principle, except that you have a matrix of panning, reverb and delay information. Just as fake if not more so. I have yet to be convinced that it is actually worth it.
Give the parent some points! This is the most sane and reasonable view of Stallman's position and actions I've seen so far. I was going to post something similar, but this is much better. Thanks Tadgh! I completely agree.
Slashdot Mirror
Exactly, and as it stands right now this is already happening.
Scientific computing has gone from exhaustive and detailed simulation to exhaustive analyses of an entire parameter space with the advent of new life science branches such as Genomics, Proteomics and all the other broadly based omics-type endeavors. This means embarrassingly parallel computation at a massive scale. At my institution we keep roughly 3000 cores humming around the clock without any difficulty, largely using legacy code.
At home I have a recording studio where the bulk of the processing is happening through plugins. Again, embarrassingly parallel workload, very easy to keep a number of cores working at 100% TODAY.
That's just two examples, but I could mention many more, and so could many of you I'm sure. I fail to see why today's software cannot make productive use of multicore architectures. Unless we're talking productivity apps, which don't use much CPU anyway...
No offense, but I can see where your handle comes from. Quit coffee, you'll be surprised.
I thought the same initially. But consider a world where one of these wireless charging technologies becomes a widely used standard. Good-bye gadgets, hello universal recharging! It's all about standards, simple ones if possible. Wireless just adds another convenience factor.
Now if we could also standardize wall outlets worldwide...
That, my friend, is a dumb idea. My Skype connection to my parents overseas, for example, is based on a P2P protocol. Giving it low QoS would badly hurt it. What I'm saying with this is you can't foresee what the next grat app is going to do under the hood, and you don't want to shoot yourself in the foot at all. Keep QoS out of the hands of most, if not all people, because if you let them choke things they will. Guaranteed!
Why are you so down on the concept? Looking for what does NOT work may be just as illuminating as our more normal way of filtering through the vast amount of sequence information. I was involved in a research effort which asked the same basic question as the one discussed here, except that we didn't dare to think of nullomers as "lethal" or "toxic". Our thoughts were more along the lines of chemically or structurally instable. We did find a large number of peptides not represented in the published databases, but the sheer number and diversity even at the tetramer level made us abandon the analysis.
Completely correct. I've done this exact calculation/database lookup a few months ago, and grep was the tool. Due to the size of the dataset, a flexible divide-and-conquer approach was needed, however, and it took a bit more than one afternoon of Ph.D. time to get it to work reliably. In the end, though, it wasn't really rocket science.
What bugs me in TFA are all the buzzwords and the focus on high level biological mechanisms. Sure, some sequences might be lethal or may cause immune reactions, but the simpler and much more feasible explanation would be that certain sequences just aren't chemically stable, at the DNA, RNA or protein level. My chemist coworker had all sorts of theories.
But to be honest, and not to be troll, I found Mac OS X to be relatively stupified in comparison to other OS's that I've used. OS X is pretty and all, but I prefer 'functional' over 'shiny' and I like to really dig into the inner workings of the OS that I use. OS X doesn't entirely appeal to my demographic for that reason, which is why it appeals to people who just want to use a computer and have it work without having to mess with it any more than they need to.
I find this a really odd mix of statements. I've beeen using versions of Unix for possibly longer than you've been around, which I think gives me a decent perspective of what makes an OS "shiny", "functional", useful or whatever else you may want of it. Currently I'm a very happy and productive OSX user. The last time I got excited about a box under my table was in the mid 90s over a decent SGI. Since then it's been boring and/or tedious, especially during the three years I used a Linux PC as my primary desktop. I love messing in the internals of an OS, but only when I want to. At other times I have better things to do. My younger colleagues seem to feel similarly, so I fail to see where you take your claims from.
As a parent of college age kids, may I suggest you might have made a few mistakes? I'm willing to bet that half of your upfront expenses could have been deferred or lowered. It does take some planning, determination and plain old "just say no", however. Life can be really good, if you want it. It's up to you to make it happen.
I'd have to agree at least in part. The higher up a guy is on the totempole the less information they utter, and replace it with buzzwords, filler and marketing fluff. Much of this interview was like this.
Still, the first page had some interest: how the two companies started to talk, customer-centered.
I find your post rather callous. While you may be right that breakthroughs don't happen without associated risks and the occasional negative or outright dangerous result, I believe we've been extremely careless during the 20th century. Your Xray example is a good one. Physicists and biologists knew fairly early on Xray radiation was ionizing, but for quite some time it didn't occur to anyone to not expose themselves or others to high doses. How hard would it be to remain a bit cautious? And maybe save a few lives and make countless other better in the process.
TFA simply advocates caution and diligent research into negative consequences of nanotech while the technology is being developed. TFA never urges abandoning anything. I agree with the author that we should keep close tabs on this stuff and watch it for long term effects.
You're correct, the iPod has these features, but they are minor items and never get in the way of accessing and playing the music, photos and video.
What I meant is that Apple demonstrates that you can resist the urge to overload and clutter the interface, onscreen as well as button count and placement.
[em]Given Apple's track record, I'd say they're the most likely candidates to figure out an elegant cell phone interface, and I'm looking forward to the iPhone for that reason alone.[/em]
You're right on the money here. Apple continue to show other makers how to not overload a music player. If they can pull off a sleek and functional phone, that would force the other manufacturers back to their drawing boards. Job's credo that you cannot design a product via focus groups still holds.
I wouldn't want to be in your shoes in twenty year's time then...
What people forget, and what the study with all its flaws doesn't touch on either is the mental energy cost of switching from paying attention to the actual task to reorganizing the workspace and then back to the task. I'm not a cognitive psychologist or whatever other discipline would have the tools to quantify task switching costs, but I have seen and felt many times how I and people I have worked with loose inordinate amounts of energy to task switching.
I'd love to see a study which addresses this and tries to quantify the productivity increase one can gain from minimizing task switching. I am convinced that in the context of using large displays this is bigger than Amdahl's law as you invoked, even though your argument is certainly valid.
Gapless playback seems to work on my old style Shuffle.
The CoverFlow view is awesome. Near instantaneous searching through your sensory channel with the most bandwidth - visual!
I sort of understand the concern too, but the reality is what makes me furious. OLGA has a ton of bad quality stuff, I agree with you on that. The official music publishers, however, also put out a ton of shit, maybe even more. Very often, sheet music for popular music is so dumbed down it's not even funny. And quite often it's not in the key it was recorded, so once the poor amateur has worked out the song, she can't even play along with the CD. That's after shelling out $3 - $5 for single song sheet music or $15 - $20 for an entire album's worth. How awful is that?
In my experience (and talking about popular music from 1960 onwards only, not jazz or classical), the only official sheet music worth the asking price is the complete Beatles book. Most everything else is a waste of your time and money. I wish OLGA and all similar sites the best of luck in their battles. They possibly do less disservice to the song writers' art than the 'official' guys do.
Total cost of ownership. The MacOSX releases really add up over the years, while you only had to pay once (but much more) for XP. Of course, you also had to buy antivirus software, muck with popup blockers and get a spyware removal tool if you were using XP, so who am I kidding?
Interesting analysis, and I have to agree to an extent. I am a Unix nerd (scientific computing) and have been an SGI customer for most of the 90s. Then commodity hardware running Linux became fast enough and had enough graphics horsepower that a switch was a nobrainer, as you say.
However, I switched away from Linux PCs on the desktop, and so have many other Unix-centered scientists, and we now use Apple computers. True, most of us like for the computer to Stay Out Of The Way, but most of us still like to be smart about computers. We just got tired of having to put the same pieces together over and over again. In today's world we can focus on getting work done, whether that be heavy duty simulations, visualization, searching databases or the web, writing up results in off the shelf office software, or prepring the next seminar in standard slideware. All at the same time, and without funny glitches. However, should Apple ever misstep we would move away in a heartbeat, unlike traditional Mac users.
On the HPC side it's much the same. We still have an SGI Altix system which can't be beat in terms of scalability of our main applications, but we have 10x as many x86-64 CPUs in a dumb cluster for the simpler 'gotta get 10,000 of these calculations done' jobs. Vendor and even Linux distro don't matter at all, we buy whatever works and comes cheap enough. Very different from the old days when we benchmarked half a dozen vendors' proprietary Unix systems for several months before settling on one, and then spent several more months in 'friendly user mode'.
The Core Duo/Solo Mac Mini and MacBooks run faster if you pair RAM sticks due to the memory architecture. This is most felt on the Minis and MacBooks (non-Pro) because they use system RAM as video RAM. From the factory all these Macs come with single RAM sticks.
Great example! Now imagine this multiplied a thousandfold or more, up and down the aisles. People will stop going there, period. See what I mean?
That's on top of the ability to forge the data in these databases. Have you ever tried to tell someone at a bank that the information they have about your credit is wrong? They'll believe the computer over you any day. The computer is always right.
That's our salvation eventually. The data will be so shitty and useless as to not matter much anymore. For all the talk about realtime and linked this and that, I have a sense that much of the data will be nearly immutable and thus unable to reflect the real world. Hence useless.
Quote: Honestly, I believe that this view of "the convergence" is as overly optimistic as the 1950's notion that by now we'd be travelling in flying cars, have robots cleaning our house*, and atomic power was going to make electricity too cheap to meter. In real life there's too much (friction? drag? entropy?) due to the sheer scale and complexity of legacy systems for things to happen the way the dreamers envision.
I was thinking the same. Look at TIA and other large scale programs. They seem to be mostly useless for their intended purpose. I think we'll end up with vast quantities of bad data from which interested parties will try to divine "insight". My local mom-and-pop corner store does better than that by simply using their eyes, ears and brains. Old tech, but works great on a local scale.
Agreed!
We need to keep in mind, too, that with the exception of classical and some jazz, all music is recorded as mostly mono tracks, processed over and over, and finally "spatialized" with pan, delay and reverb. Whether stereo or surround, it's all fake room information, no matter how good the underlying technology is. Like you, I think the current music surround craze is totally manufactured.
I completey agree with you, and would go even further: Most of today's music doesn't even have real LEFT and RIGHT information either. Most instrument and vocal tracks start out in MONO, and the left/right information is added via judicious application of pan and reverb and maybe delay. In other words, it's all fake. This technique does give us a pleasing separation of instruments and voices, which is great, and it gives a sense of space, which is necessary. But true spatial information it is not.
Now throw in today's craze to mix as surround sound. Same general principle, except that you have a matrix of panning, reverb and delay information. Just as fake if not more so. I have yet to be convinced that it is actually worth it.
Give the parent some points!
This is the most sane and reasonable view of Stallman's position and actions I've seen so far. I was going to post something similar, but this is much better. Thanks Tadgh! I completely agree.