Because if you are actually using that many peripherals with your laptop at the same time..
More than 2? You are arguing that 2 USB ports is enough.
External HD + Mouse.. so no other peripherals at all.. no connecting your MP3 player or synching your phone.. no thumb drive.. none of those other devices.. zero.. oh, I should disconnect one of those two? really?
When recording from any unfiltered analog source to a quantized digital format, frequencies above the sampling rate (which always exist in the unfiltered) become aliasing artifacts, similar to the moire sampling noise seen in the imaging world. The simple solution in the graphics world is to of course photograph in a higher resolution and then down-sample to the desired resolution. Going over 44.1khz in the audio field is the same general solution to this sampling problem.
The problem gets worse with small re-sampling adjustments that many computer users are doing without knowing it. Many motherboards come with 48khz DAC's so playing 44.1khz audio data (pretty much all audio data) is actually far less than ideal, "blurring out" the dynamics of the higher frequencies in the original.
- hard drives, okay for low speed (that includes USB flash drives), but there's a FireWire 800 port built-in that's much faster
FireWire is irrelevant if all my devices are USB.
- sound cards, built-in inputs and outputs
..and do those offer 24-bit 192khz with superb SNR like my E-MU 404 USB soundboard? How about a MIDI port.. oh it doesnt even have that, so I need a USB midi controller anyways? Yeah...
- headsets, yes some are USB but most are still using 1/8" jacks
1/8" Wireless? Yeah.. not.
webcams, built-in
..so not as good as the one I already have thats USB?
Whats wrong with two at the corner, one on each edge? This design would seem to be the most accommodating. Better still, both back corners could be this way, so that there are 2 USB's on the back (opposite sides) as well as 1 USB for each of the left and right sides.
If I was designing a laptop, thats what I would do. Maybe nobody but me sees the value in 4 USB ports in an age where just about every peripheral can be gotten with a USB interface (hard drives, cd/dvd, mouse, sound cards, headsets, compact flash readers, webcams, ethernet ports, phones and MP3 players,...)
The Internet is only one such network. There were more than a few such networks in the 70's, 80's, and 90's (Telenet and Tymnet to name a few)
The fact that the Internet "won" is in no small part due to government cash building out the initial infrastructure making it hard to compete against. It wasnt because there was no competition, it was because they could not compete.
Give her a gift card for a spa or other "nice" thing to do for the day. She will (A) love you for it, (B) never need to know that you had a kiddie porn drive, or (C) that you baked said kiddie porn drive in the oven while downloading midget porn as a replacement.
That's right, there is no such thing as free will or consciousness.
Ah, the old reaching-for-ill-defined-hocus-pocus rebuttal. You might as well just invoke "the mind is a paranormal phenomenon" or "what about gods influence"
This idea that the mind is somehow more than physical, that it somehow does not obey the laws of physics... is completely unsubstantiated bullshit.
Thats just not good enough. Its an excuse to moan about Apple/Microsoft, rather than an excuse not to do business with them.
The author could make the source available as GPL without distributing it through an app store as GPL-licensed. Thats all there is to it. Not wanting to is the same as not wanting to sell your application on the app store. The author losing NOTHING by using a separate license for the app store distribution. Refusing to do so it cry-baby shit, unless of course the person doesnt have the rights to do it.. in which case they also do not have the rights to sell other peoples GPLv3 works on a locked down system.
Microsoft is basically a third party here.
This has nothing to do with the party-status of Apple/Microsoft. The GPLv3 is most definitely not compatible with a locked down ecosystem. Either waive the GPLv3 license from the app stores distribution (if you have the rights to do so) or dont cry about it (because you dont have the right as defined by the license of the code portions that arent yours)
Its as if you think you get to take from the GPL without honoring the full terms of the license yourself while calling Apple/Microsoft evil because they wont allow you to violate the license terms you specified.
Here's an idea.. honor the fucking license and don't do business with Microsoft or Apple, or don't use the license. If you can't avoid using the license, its because its not your fucking work.
BSD puts burden on the distributor. You have to keep the copyright notice.
No, BSD puts a burden on the author that uses BSD-licensed code to directly include said stuff within the distribution.
Redistributions in binary form must reproduce the above copyright notice, this list
of conditions and the following disclaimer in the documentation and/or other materials
provided with the distribution.
Why are Anonymous Cowards always so often plainly wrong?
But why doesnt the author simply re-license it as something other than GPLv3 for the App Stores version?
Light bulb! Its because he doesnt own all the copyrights. Thats the nature of GPL, after all. One person building on another, with no individual having any rights to change the terms of the license. The person submitting the App can show that he is claiming full ownership by simply changing the license. If they can't do that, then why does Apple or Microsoft want to deal with them?
I do know that pointing to Apple and saying "they do it too" is a logical fallacy for justifying evil.
You seem to be forgetting precisely why Apple does it too. Its because the authors of GPL software went after them. It isnt that Apple didn't want GPL software in the App Store, its that they got more grief from the copyright holders than it was worth.
This idea that Apple is "evil" because they are disallowing GPL software is disingenuous at best. It began with GPL authors being the ones that disallowed their applications on the app store, so Apple says "We want to deal with only one entity, and entity that owns all the rights to the program. That GPL license just causes us grief. If you have the right to relicense it, then we can do business, otherwise we can't."
Even BSD is restrictive, and thats way more open than GPL v2/v3.
The sad fact is that BSD is losing vs GPL precisely because BSD licensed source code can be used in GPL software, but not vise-versa. That right there is proof that the GPL is anything but "open," that it is in fact "viral" as has been pointed out numerous times.
That sort of thing would only add a logarithmic term to the runtime complexity (see logarithmic methods in N-body gravity simulations), so can simply be ignored for the purposes of this conversation.
To boot, I think most people are ignoring even higher level methods than simple neuron sims. We are talking about a rather sloppy "device", after all. The specific state of a single neuron just doesnt mean much (the physical effects of motion guarantee that.)
The human brain is not anything like the computers we build today, except to say that they can both "compute", to varying degrees.
You are arguing that because the hardware of a computer is nothing like the hardware of a human that this supports your point, but it doesnt. It is only if the software running on the computer was nothing like the software of a human that you would have a point.
All of todays CPU's are universal turing machine, and that fact negates every single hardware argument. We've heard this all before, from people saying that current computers arent fast enough to simulate the human brain (your arstechnica link makes this claim in a round-about way, but fails to realize that the simulation does not have the be real-time) to this tripe about Von Neumann architectures not "working like a brain" (why does the hardware itself have to work like the brain?)
Basically, you are parroting the fallacious "pop sci" arguments that have no bearing on what real computer scientists know. Now, for sure this Watson wasn't "thinking like a human", but that is not the crap that you just spewed at us. You went off on the hardware bullshit and further linked to the speed bullshit.
Most of the OS files are read-once/write-never (if what gets loaded gets written, its to the swap file), so is simply not a candidate for any worthwhile caching algorithm.
The problem is that there is no "the operator is waiting" signal that the drive will ever get. Computers do plenty of disk I/O while we are not waiting and there is just no way for that to be measured in drive firmware.
The best the drive could do is reduce the time the computer itself waits.. but that is not representative of the human's pain (that wants the OS to boot fast, and when he runs photoshop however rarely he wants it to open instantly.. but doesn't care that a background autosave takes a second or two since he NEVER feels it)
The market for consumer 10000 RPM drives is almost completely eradicated. They are down to about $0.50/GB now at best, but thats also the awkward capacities like 600GB.
That capacity is awkward because if you are throwing $300 at performance storage from the consumer space, then you are a fool to choose any single platter vs a 160GB SSD like the OCZ Vertex 2 (over 2 times as much bandwidth and 180 times as many IOPS vs any 10000 RPM drive.)
Combine this with the pressure from the 2 TB 7200 RPM platters which have the same or better throughput and only marginally worse IOPS than the 10000 RMP's, and doing it for only $100, and well... you see the point that you can RAID several of these 2TB 7200's for less than a 10000 RPM and get much better performance (which is why I said "single platter" in the previous paragraph)
Basically I'm saying that the 10000 RPM drives like the VelociRaptor were attacked from both sides. There is simply no room in between the cheap-but-similarly-performing 2TB platters (at $0.05/GB) and the SSD's (at $1.70/GB)
ARM is also fabless, but the difference is that AMD contracts out for the production of chips whereas ARM simply sells chip designs.
This is a significant distinction, as AMD still has contractual control over the baking, whereas ARM couldnt give a shit about anything other than selling their IP.
Secondly..
AMD owns the ATI fabs and continues to produce GPU's from them.
Those old VelociRaptors arent even competitive with a modern 2TB 7,200K RPM consumer drive like the Caviar Black in performance (which is less than $200)
Even ignoring SSD's, that class of 10000 RPM drive has been eliminated from the performance market by the higher capacity 7200 RPM drives. The main issue is that while 10K RPM still gives better seek times, that same RPM also keeps them from using the highest drive densities. 7200 RPM at a higher density beats 10000 RPM at a lower density on raw throughput.
So what we end up with is that the Raptor's having no market any longer. For throughput they are no longer competitive with larger and cheaper drives, and for IOPS they are simply a joke compared to even thumb drives. There is no market for consumer-grade 10K RPM like there was in the past.
The only 10K+ RPM drives that are still successful are enterprise-class, and those aren't cheaper than SSD's. Some enterprise drives can be had for ~$1/GB but so too some SSD's can be had for that, and most enterprise-class drives will run you ~$2/GB.
The shortcoming of SSD's is that they carry an enterprise price tag but do not carry the same enterprise-level guarantees, but that is offset by the significant performance advantages that they do offer.
Essentially, you are reaching for yesteryears trendy performance geek stuff and trying to apply it to todays performance geek stuff. Performance geeks have been using SSD's for several years now, and thats simply not going to swing back towards platters... ever.
Slashdot Mirror
Because if you are actually using that many peripherals with your laptop at the same time..
More than 2? You are arguing that 2 USB ports is enough. .. so no other peripherals at all.. no connecting your MP3 player or synching your phone.. no thumb drive.. none of those other devices.. zero.. oh, I should disconnect one of those two? really?
External HD + Mouse
People may also not be aware that most CD players use 1-bit DAC's...
99.9% of music sold is compressed into the top half of the spectrum
Are you aware that the "top half of the spectrum" as you call it would be a loss of only a single bit of resolution? 15-bit instead of 16-bit?
Just saying...
When recording from any unfiltered analog source to a quantized digital format, frequencies above the sampling rate (which always exist in the unfiltered) become aliasing artifacts, similar to the moire sampling noise seen in the imaging world. The simple solution in the graphics world is to of course photograph in a higher resolution and then down-sample to the desired resolution. Going over 44.1khz in the audio field is the same general solution to this sampling problem.
The problem gets worse with small re-sampling adjustments that many computer users are doing without knowing it. Many motherboards come with 48khz DAC's so playing 44.1khz audio data (pretty much all audio data) is actually far less than ideal, "blurring out" the dynamics of the higher frequencies in the original.
- hard drives, okay for low speed (that includes USB flash drives), but there's a FireWire 800 port built-in that's much faster
FireWire is irrelevant if all my devices are USB.
- sound cards, built-in inputs and outputs
- headsets, yes some are USB but most are still using 1/8" jacks
1/8" Wireless? Yeah.. not.
webcams, built-in
ethernet ports, there's one built-in
So only one? Yeah.
Whats wrong with two at the corner, one on each edge? This design would seem to be the most accommodating. Better still, both back corners could be this way, so that there are 2 USB's on the back (opposite sides) as well as 1 USB for each of the left and right sides.
If I was designing a laptop, thats what I would do. Maybe nobody but me sees the value in 4 USB ports in an age where just about every peripheral can be gotten with a USB interface (hard drives, cd/dvd, mouse, sound cards, headsets, compact flash readers, webcams, ethernet ports, phones and MP3 players,...)
The Internet is only one such network. There were more than a few such networks in the 70's, 80's, and 90's (Telenet and Tymnet to name a few)
The fact that the Internet "won" is in no small part due to government cash building out the initial infrastructure making it hard to compete against. It wasnt because there was no competition, it was because they could not compete.
Perhaps we should also ask re-captured prisoners why they tried to escape.
Give her a gift card for a spa or other "nice" thing to do for the day. She will (A) love you for it, (B) never need to know that you had a kiddie porn drive, or (C) that you baked said kiddie porn drive in the oven while downloading midget porn as a replacement.
That's right, there is no such thing as free will or consciousness.
Ah, the old reaching-for-ill-defined-hocus-pocus rebuttal. You might as well just invoke "the mind is a paranormal phenomenon" or "what about gods influence"
This idea that the mind is somehow more than physical, that it somehow does not obey the laws of physics... is completely unsubstantiated bullshit.
Because he doesn't want to?
Thats just not good enough. Its an excuse to moan about Apple/Microsoft, rather than an excuse not to do business with them.
The author could make the source available as GPL without distributing it through an app store as GPL-licensed. Thats all there is to it. Not wanting to is the same as not wanting to sell your application on the app store. The author losing NOTHING by using a separate license for the app store distribution. Refusing to do so it cry-baby shit, unless of course the person doesnt have the rights to do it.. in which case they also do not have the rights to sell other peoples GPLv3 works on a locked down system.
Microsoft is basically a third party here.
This has nothing to do with the party-status of Apple/Microsoft. The GPLv3 is most definitely not compatible with a locked down ecosystem. Either waive the GPLv3 license from the app stores distribution (if you have the rights to do so) or dont cry about it (because you dont have the right as defined by the license of the code portions that arent yours)
Its as if you think you get to take from the GPL without honoring the full terms of the license yourself while calling Apple/Microsoft evil because they wont allow you to violate the license terms you specified.
Here's an idea.. honor the fucking license and don't do business with Microsoft or Apple, or don't use the license. If you can't avoid using the license, its because its not your fucking work.
BSD puts burden on the distributor. You have to keep the copyright notice.
No, BSD puts a burden on the author that uses BSD-licensed code to directly include said stuff within the distribution.
Redistributions in binary form must reproduce the above copyright notice, this list
of conditions and the following disclaimer in the documentation and/or other materials
provided with the distribution.
Why are Anonymous Cowards always so often plainly wrong?
No, but I am more inclined to break the law for minor gains than Apple is. There is no shortage of non-GPL apps on their store.
But why doesnt the author simply re-license it as something other than GPLv3 for the App Stores version?
Light bulb! Its because he doesnt own all the copyrights. Thats the nature of GPL, after all. One person building on another, with no individual having any rights to change the terms of the license. The person submitting the App can show that he is claiming full ownership by simply changing the license. If they can't do that, then why does Apple or Microsoft want to deal with them?
I do know that pointing to Apple and saying "they do it too" is a logical fallacy for justifying evil.
You seem to be forgetting precisely why Apple does it too. Its because the authors of GPL software went after them. It isnt that Apple didn't want GPL software in the App Store, its that they got more grief from the copyright holders than it was worth.
This idea that Apple is "evil" because they are disallowing GPL software is disingenuous at best. It began with GPL authors being the ones that disallowed their applications on the app store, so Apple says "We want to deal with only one entity, and entity that owns all the rights to the program. That GPL license just causes us grief. If you have the right to relicense it, then we can do business, otherwise we can't."
Man I 100% agree with you.
Even BSD is restrictive, and thats way more open than GPL v2/v3.
The sad fact is that BSD is losing vs GPL precisely because BSD licensed source code can be used in GPL software, but not vise-versa. That right there is proof that the GPL is anything but "open," that it is in fact "viral" as has been pointed out numerous times.
Watson is still just following a fixed set of rules
So are you. We call it physics.
That sort of thing would only add a logarithmic term to the runtime complexity (see logarithmic methods in N-body gravity simulations), so can simply be ignored for the purposes of this conversation.
To boot, I think most people are ignoring even higher level methods than simple neuron sims. We are talking about a rather sloppy "device", after all. The specific state of a single neuron just doesnt mean much (the physical effects of motion guarantee that.)
The human brain is not anything like the computers we build today, except to say that they can both "compute", to varying degrees.
You are arguing that because the hardware of a computer is nothing like the hardware of a human that this supports your point, but it doesnt. It is only if the software running on the computer was nothing like the software of a human that you would have a point.
All of todays CPU's are universal turing machine, and that fact negates every single hardware argument. We've heard this all before, from people saying that current computers arent fast enough to simulate the human brain (your arstechnica link makes this claim in a round-about way, but fails to realize that the simulation does not have the be real-time) to this tripe about Von Neumann architectures not "working like a brain" (why does the hardware itself have to work like the brain?)
Basically, you are parroting the fallacious "pop sci" arguments that have no bearing on what real computer scientists know. Now, for sure this Watson wasn't "thinking like a human", but that is not the crap that you just spewed at us. You went off on the hardware bullshit and further linked to the speed bullshit.
AutoPagerize for Opera should work. Havent tried it myself. It claims to only inline the next page when you reach the end of the article.
Bullshit.
Most of the OS files are read-once/write-never (if what gets loaded gets written, its to the swap file), so is simply not a candidate for any worthwhile caching algorithm.
The problem is that there is no "the operator is waiting" signal that the drive will ever get. Computers do plenty of disk I/O while we are not waiting and there is just no way for that to be measured in drive firmware.
The best the drive could do is reduce the time the computer itself waits.. but that is not representative of the human's pain (that wants the OS to boot fast, and when he runs photoshop however rarely he wants it to open instantly.. but doesn't care that a background autosave takes a second or two since he NEVER feels it)
The market for consumer 10000 RPM drives is almost completely eradicated. They are down to about $0.50/GB now at best, but thats also the awkward capacities like 600GB.
That capacity is awkward because if you are throwing $300 at performance storage from the consumer space, then you are a fool to choose any single platter vs a 160GB SSD like the OCZ Vertex 2 (over 2 times as much bandwidth and 180 times as many IOPS vs any 10000 RPM drive.)
Combine this with the pressure from the 2 TB 7200 RPM platters which have the same or better throughput and only marginally worse IOPS than the 10000 RMP's, and doing it for only $100, and well... you see the point that you can RAID several of these 2TB 7200's for less than a 10000 RPM and get much better performance (which is why I said "single platter" in the previous paragraph)
Basically I'm saying that the 10000 RPM drives like the VelociRaptor were attacked from both sides. There is simply no room in between the cheap-but-similarly-performing 2TB platters (at $0.05/GB) and the SSD's (at $1.70/GB)
Well that is sort of misleading.
First of all..
ARM is also fabless, but the difference is that AMD contracts out for the production of chips whereas ARM simply sells chip designs.
This is a significant distinction, as AMD still has contractual control over the baking, whereas ARM couldnt give a shit about anything other than selling their IP.
Secondly..
AMD owns the ATI fabs and continues to produce GPU's from them.
What sort of failure?
If its not simply using up the write limits (which requires extreme usage), then it sounds like something else is amiss.
Those old VelociRaptors arent even competitive with a modern 2TB 7,200K RPM consumer drive like the Caviar Black in performance (which is less than $200)
Even ignoring SSD's, that class of 10000 RPM drive has been eliminated from the performance market by the higher capacity 7200 RPM drives. The main issue is that while 10K RPM still gives better seek times, that same RPM also keeps them from using the highest drive densities. 7200 RPM at a higher density beats 10000 RPM at a lower density on raw throughput.
So what we end up with is that the Raptor's having no market any longer. For throughput they are no longer competitive with larger and cheaper drives, and for IOPS they are simply a joke compared to even thumb drives. There is no market for consumer-grade 10K RPM like there was in the past.
The only 10K+ RPM drives that are still successful are enterprise-class, and those aren't cheaper than SSD's. Some enterprise drives can be had for ~$1/GB but so too some SSD's can be had for that, and most enterprise-class drives will run you ~$2/GB.
The shortcoming of SSD's is that they carry an enterprise price tag but do not carry the same enterprise-level guarantees, but that is offset by the significant performance advantages that they do offer.
Essentially, you are reaching for yesteryears trendy performance geek stuff and trying to apply it to todays performance geek stuff. Performance geeks have been using SSD's for several years now, and thats simply not going to swing back towards platters... ever.