If they had machines doing that work, then those people could be doing something else, effectively doubling their contribution. At least, if China were a REAL communist country, that logic would hold.
CRAPS performance depends in a big way on what you're doing with it. For anything math intensive, newer x86s tend to be better per clock when the vector unit is used. For I/O bound tasks and those with more string-oriented functions [ie, less math], CRAPS can [and I would think they usually do] perform better.
GPL v3 is really no different to GPL v2, and I think developers realise this, both developers coding for money, and those coding for fun. It's not likely to change any business practices that aren't already an abuse of the GPL, and it's not likely to change the mind of developers who don't want to get shafted by some bottom dwelling corporate rip-off artists. So where, then, is your point?
Re:Interesting -- but viable?
on
Photosynth Demo
·
· Score: 0
"I read this has enormous space requirements *and* computations associated with it."
I wouldn't think so, at least, the computational requirements. I had what I believe was a similar algorithm in some software I worked on a while back, and got fantastic performance, without hardware acceleration. When you get clever with OpenGL.. well..
We can think about/entire/ three dimensional objects, though. There's replacing the mouse, and then there's more. For example, you could use such a computer to hold thoughts, you know when it all zips by so fast? Like when you're trying to paint and you get part of the picture here and part there... it could record all the parts and let you modify large sections of the image in real time. Of course, something the size of an aspirin won't allow you to capture an entire image from the visual cortex or read it back, but I'm not certain external brainwave monitors will be able to get that sort of resolution.
I had some good experiences with Diaryland a while back. This is making me consider moving back there. I can't think of one thing LJ does better than DL.
I'd like to think compilers should be smart enough to assess data dependencies and space those instructions out- we've always had to do this anyway, at least since pipelined processors hit the market, but loops still aren't cascaded properly. An example is a loop that calculates a sum of products- the add instruction must wait for the multiplication instruction to finish, when in fact the processor could be doing a heap of multiplications, and using associativity to cut down dataflow problems in the add stage. Spreading the dataflow graph out as much as possible at compile time also helps with cache coherency between many processors.
I think a program compiled that way would need hardware that will understand that the data dependencies are spread out so that it can distribute instructions among the processors, although the distribution could be very simple if the dependencies could be spread out significantly- instructions could almost be distributed like dealing cards. It's a much finer granularity than threading, but I think more applications suit this sort of parallelism.
Another barrier to parallel programming is accessing [for read] data that should be global to all threads. You can do it by passing pointers to state [messy], using globals [dangerous, obese] or by copying all the data onto the stack of the thread [slow]. Threads need to really share address space- use the same stack and everything, IMHO.
I could be wrong, but I don't see how Apple are bringing gender into this. I'd like you to point out one piece of evidence that supports your claim of this misogynistic Apple. Otherwise, please take your cargo cult elsewhere.
Side to side? Try operating more than one button at once when the buttons are all next to each other. On the SNES controller, you can hold the Y button and still operate the B button or vice versa.
Personally, I think most modern controllers are too small. It started with the PSX controllers and Microsoft continued it. I can't chose between having an analog stick and having a decent sized controller though.
Modern Nokias have the same feature- the unlock key is uncovered. Every time my father sits down, the mp3 player starts. That and phones where emergency dialing numbers are dialable when the phone is locked- I've heard "Police Fire Ambulance. Hello?" come from my pants while at work. Rather unnerving.
This of course stems from a more common problem: phones don't use open source operating systems, because that'd make these stupid little things very easy to fix. Either give me that or give me a phone that doesn't need its own operating system, thankyou.
I heard good things about the wireless support in Feisty too, and when some friends tried it out they said it still hasn't been fixed. And it's so strange, because other distributions work fine.
While your point on gaming is a good point [Windows games typically taking an extra 2-3 minutes to set up on a Linux/Unix system], I don't buy the hardware thing. I've never had a hardware support problem under Linux or Unix like the sort of issues I've had under Windows, however, I know someone who has [with a certain wifi card last year]. But, even when most hardware is designed to be used with a Windows machine, it's still usually easier to use and better supported under Linux.
Selling the machines at a loss does not mean they lose more money the more they sell. The hardware sells at a profit, while it's one-off stuff they are still paying for, like R&D, design, and testing, not to mention reasonably fixed price stuff like marketing. If they were to sell enough Xsoxes they would actually make up the loss.
Heh, that's not quite what I meant about the universe curving back on itself, I meant more like in a hypersphere shape. Energy [and by that I mean mass] curves the universe. It's like this: if you take a piece of paper and bend it in enough places in all the right directions, the edges of the paper will meet. That's effectively what gravity is: where there is mass or energy, the universe is bent. Of course, we can't see it bending, but we can measure it. This deviation from Euclidean geometry is very slight, but on a large scale, it means that distance would essentially be finite and the universe closed- if you travel too far, you come back to where you started. It's a given on any Riemann surface with plenty of points of positive curvature and none of negative curvature.
In a closed universe such as ours, then, how can it be possible for something to be "outside" the universe? For example, what direction would it be in? If you go that way you will end up where you began.
If other galaxies and far away stars are observably affected by your megaverse, I put it that the megaverse would be observable. Since we see the movement of galaxies millions of years after it happened, it would take the same ammount of time for the gravity to also reach us. But, it would be possible that the acceleration is very small, and therefore undetectable, but given enough time the acceleration integrates into the velocities we see galaxies moving at. [ie. possible.]
Gravity pushing actually does make sense, especially if you think in terms of quanta. For momentum to be imparted via gravity, that momentum needs to be sent out as a particle, and must carry with it that momentum. And actually, I think that itself explains why galaxies appear red shifted. Quite simply, photon wave functions grow in standard distribution [as a Green's Function solution of a diffusion equation where the impulse is the mean path of the photon], and because the effect is essentially spread out, its effect is diminished by the curvature itself.
Gravity is a lot like an ideal fluid. Gravity is a conservative vector field, and energy changes the field, and the field changes the energy, a lot like a liquid flowing. But what you're suggesting requires you break the conservation, including having gravity sinks. The big thing you have standing in your way is harmonising these waves. Gravity does, at least on the nanometre to millions of light year scale, act very predictably. Why it would differ from that is something you're going to have to work out, so you can find a method to test it. That it's not observable probably means that it's not relavant.
Speed of light being definite doesn't mean that there is a maximum acceleration. You can keep adding momentum to a body and it will keep accelerating proportionally, and yet never reach the speed of light. Crazy, huh?:)
By the way, you can throw a ball ahead of you if you're on a rocket traveling close to the speed of light. In fact, if you're traveling close to the speed of light, you won't even know it until you look out the window. Relativistic physics says that speed is relative, it doesn't matter how fast your rocket is going, if you're not moving relative to the rocket, things will be essentially the same as normal. If you're on a rocket and I'm not, you could say that I am actually moving at close to the speed of light, and therefore I can't throw a ball in the direction you were coming from.
I think you will definitely have to write more [I didn't understand a lot of that, sorry]. I've had a number of crazy universe theories myself, though I've mostly debunked them, I'm always looking. I hope you do push your theory further and ask yourself what would happen if... and then look for evidence.
You're missing the multi user point. The great thing about having more than one user is you can have a program run as it's own user, without write access to anything [but its home directory], and without read access to anything personal or private. If permissions were as complicated as you say, installers would set them, and people wouldn't look and correct them.
But then how do programs write to the home directory of the calling user? I'd like to see standardised save-as boxes and such that handle passing that sort of permission, and any other attempt at writing outside the permissions of that program should be interrupted and a query sent to the user. And then, to simplify things, it would be nice to pass EROS style capabilities to programs permanently, so they don't need to explicitly request access to other things they may need.
Though, once you start restricting things like that, you need a whole new model for IPC and such... and I don't think we're ready for that yet.
You're definitely a thinker, but let me explain why I think your Megaverse is unlikely to be our universe.
The universe we can see now contains enough matter to completely curve the universe back on itself [which is to say, it must curve back on itself in at least one dimension]. If the remnants from another big bang were to come in contact with our universe, would they do so from inside or outside the universe? In any situation, there would be enough energy in any one area to severely affect the net curvature and effectively make the entire previous universe smaller.
Though I agree with you, the big bang theory is too problematic, IMHO because a tiny universe should have a Shwarzchild radius, and consequently shouldn't grow.
Pro Tools are not very Pro, somehow. Every time I'm stuck using them I wonder how people manage.
I'm pretty sure there is no available solution for people who want to use more than sixteen VSTi/DXI soft synths at a time. Despite its shortcomings, FL Studio scales easily to 300+ simultaneous synths over 64 main effect channels with eight effects each. If you want to do more than sampling on Linux, you're pretty much limited to a couple of synths.
Slashdot Mirror
If they had machines doing that work, then those people could be doing something else, effectively doubling their contribution. At least, if China were a REAL communist country, that logic would hold.
CRAPS performance depends in a big way on what you're doing with it. For anything math intensive, newer x86s tend to be better per clock when the vector unit is used. For I/O bound tasks and those with more string-oriented functions [ie, less math], CRAPS can [and I would think they usually do] perform better.
Amen. Or, you can write a more contextual one in five lines of [insert favourite scripting language here].
Sounds like what you need is actually an obstack. These come with glibc.
Do we really want to explore the largest Hemorrhoid in the solar system?
Because it's a great excuse to spread some FUD?
GPL v3 is really no different to GPL v2, and I think developers realise this, both developers coding for money, and those coding for fun. It's not likely to change any business practices that aren't already an abuse of the GPL, and it's not likely to change the mind of developers who don't want to get shafted by some bottom dwelling corporate rip-off artists. So where, then, is your point?
Sorry, I didn't know that.
Well, it seems a bit strange to say Microsoft are innovating then, doesn't it?
This was an acquisition ;)
"I read this has enormous space requirements *and* computations associated with it."
I wouldn't think so, at least, the computational requirements. I had what I believe was a similar algorithm in some software I worked on a while back, and got fantastic performance, without hardware acceleration. When you get clever with OpenGL.. well..
We can think about /entire/ three dimensional objects, though. There's replacing the mouse, and then there's more. For example, you could use such a computer to hold thoughts, you know when it all zips by so fast? Like when you're trying to paint and you get part of the picture here and part there... it could record all the parts and let you modify large sections of the image in real time. Of course, something the size of an aspirin won't allow you to capture an entire image from the visual cortex or read it back, but I'm not certain external brainwave monitors will be able to get that sort of resolution.
I had some good experiences with Diaryland a while back. This is making me consider moving back there. I can't think of one thing LJ does better than DL.
I'd like to think compilers should be smart enough to assess data dependencies and space those instructions out- we've always had to do this anyway, at least since pipelined processors hit the market, but loops still aren't cascaded properly. An example is a loop that calculates a sum of products- the add instruction must wait for the multiplication instruction to finish, when in fact the processor could be doing a heap of multiplications, and using associativity to cut down dataflow problems in the add stage. Spreading the dataflow graph out as much as possible at compile time also helps with cache coherency between many processors.
I think a program compiled that way would need hardware that will understand that the data dependencies are spread out so that it can distribute instructions among the processors, although the distribution could be very simple if the dependencies could be spread out significantly- instructions could almost be distributed like dealing cards. It's a much finer granularity than threading, but I think more applications suit this sort of parallelism.
Another barrier to parallel programming is accessing [for read] data that should be global to all threads. You can do it by passing pointers to state [messy], using globals [dangerous, obese] or by copying all the data onto the stack of the thread [slow]. Threads need to really share address space- use the same stack and everything, IMHO.
Didn't you already say this once?
I could be wrong, but I don't see how Apple are bringing gender into this. I'd like you to point out one piece of evidence that supports your claim of this misogynistic Apple. Otherwise, please take your cargo cult elsewhere.
Side to side? Try operating more than one button at once when the buttons are all next to each other. On the SNES controller, you can hold the Y button and still operate the B button or vice versa.
Personally, I think most modern controllers are too small. It started with the PSX controllers and Microsoft continued it. I can't chose between having an analog stick and having a decent sized controller though.
Modern Nokias have the same feature- the unlock key is uncovered. Every time my father sits down, the mp3 player starts. That and phones where emergency dialing numbers are dialable when the phone is locked- I've heard "Police Fire Ambulance. Hello?" come from my pants while at work. Rather unnerving. This of course stems from a more common problem: phones don't use open source operating systems, because that'd make these stupid little things very easy to fix. Either give me that or give me a phone that doesn't need its own operating system, thankyou.
I heard good things about the wireless support in Feisty too, and when some friends tried it out they said it still hasn't been fixed. And it's so strange, because other distributions work fine.
While your point on gaming is a good point [Windows games typically taking an extra 2-3 minutes to set up on a Linux/Unix system], I don't buy the hardware thing. I've never had a hardware support problem under Linux or Unix like the sort of issues I've had under Windows, however, I know someone who has [with a certain wifi card last year]. But, even when most hardware is designed to be used with a Windows machine, it's still usually easier to use and better supported under Linux.
Selling the machines at a loss does not mean they lose more money the more they sell. The hardware sells at a profit, while it's one-off stuff they are still paying for, like R&D, design, and testing, not to mention reasonably fixed price stuff like marketing. If they were to sell enough Xsoxes they would actually make up the loss.
Firefox would be far better to store its cache in memory and let the operating system handle paging.
On the contrary- wasted is the I/O resources used to fill up that memory when it could be doing something that is actually useful.
Heh, that's not quite what I meant about the universe curving back on itself, I meant more like in a hypersphere shape. Energy [and by that I mean mass] curves the universe. It's like this: if you take a piece of paper and bend it in enough places in all the right directions, the edges of the paper will meet. That's effectively what gravity is: where there is mass or energy, the universe is bent. Of course, we can't see it bending, but we can measure it. This deviation from Euclidean geometry is very slight, but on a large scale, it means that distance would essentially be finite and the universe closed- if you travel too far, you come back to where you started. It's a given on any Riemann surface with plenty of points of positive curvature and none of negative curvature.
:)
In a closed universe such as ours, then, how can it be possible for something to be "outside" the universe? For example, what direction would it be in? If you go that way you will end up where you began.
If other galaxies and far away stars are observably affected by your megaverse, I put it that the megaverse would be observable. Since we see the movement of galaxies millions of years after it happened, it would take the same ammount of time for the gravity to also reach us. But, it would be possible that the acceleration is very small, and therefore undetectable, but given enough time the acceleration integrates into the velocities we see galaxies moving at. [ie. possible.]
Gravity pushing actually does make sense, especially if you think in terms of quanta. For momentum to be imparted via gravity, that momentum needs to be sent out as a particle, and must carry with it that momentum. And actually, I think that itself explains why galaxies appear red shifted. Quite simply, photon wave functions grow in standard distribution [as a Green's Function solution of a diffusion equation where the impulse is the mean path of the photon], and because the effect is essentially spread out, its effect is diminished by the curvature itself.
Gravity is a lot like an ideal fluid. Gravity is a conservative vector field, and energy changes the field, and the field changes the energy, a lot like a liquid flowing. But what you're suggesting requires you break the conservation, including having gravity sinks. The big thing you have standing in your way is harmonising these waves. Gravity does, at least on the nanometre to millions of light year scale, act very predictably. Why it would differ from that is something you're going to have to work out, so you can find a method to test it. That it's not observable probably means that it's not relavant.
Speed of light being definite doesn't mean that there is a maximum acceleration. You can keep adding momentum to a body and it will keep accelerating proportionally, and yet never reach the speed of light. Crazy, huh?
By the way, you can throw a ball ahead of you if you're on a rocket traveling close to the speed of light. In fact, if you're traveling close to the speed of light, you won't even know it until you look out the window. Relativistic physics says that speed is relative, it doesn't matter how fast your rocket is going, if you're not moving relative to the rocket, things will be essentially the same as normal. If you're on a rocket and I'm not, you could say that I am actually moving at close to the speed of light, and therefore I can't throw a ball in the direction you were coming from.
I think you will definitely have to write more [I didn't understand a lot of that, sorry]. I've had a number of crazy universe theories myself, though I've mostly debunked them, I'm always looking. I hope you do push your theory further and ask yourself what would happen if... and then look for evidence.
You're missing the multi user point. The great thing about having more than one user is you can have a program run as it's own user, without write access to anything [but its home directory], and without read access to anything personal or private. If permissions were as complicated as you say, installers would set them, and people wouldn't look and correct them.
But then how do programs write to the home directory of the calling user? I'd like to see standardised save-as boxes and such that handle passing that sort of permission, and any other attempt at writing outside the permissions of that program should be interrupted and a query sent to the user. And then, to simplify things, it would be nice to pass EROS style capabilities to programs permanently, so they don't need to explicitly request access to other things they may need.
Though, once you start restricting things like that, you need a whole new model for IPC and such... and I don't think we're ready for that yet.
You're definitely a thinker, but let me explain why I think your Megaverse is unlikely to be our universe. The universe we can see now contains enough matter to completely curve the universe back on itself [which is to say, it must curve back on itself in at least one dimension]. If the remnants from another big bang were to come in contact with our universe, would they do so from inside or outside the universe? In any situation, there would be enough energy in any one area to severely affect the net curvature and effectively make the entire previous universe smaller. Though I agree with you, the big bang theory is too problematic, IMHO because a tiny universe should have a Shwarzchild radius, and consequently shouldn't grow.
Pro Tools are not very Pro, somehow. Every time I'm stuck using them I wonder how people manage. I'm pretty sure there is no available solution for people who want to use more than sixteen VSTi/DXI soft synths at a time. Despite its shortcomings, FL Studio scales easily to 300+ simultaneous synths over 64 main effect channels with eight effects each. If you want to do more than sampling on Linux, you're pretty much limited to a couple of synths.