there are lots of useful computations that are more flops-intensive (relative to memory footprint) than dot-products. matmul, fft, almost anything montecarlo, etc.
this lawyer doesn't really understand programming - even his understanding of copyright is only enough to make him dangerous. he states, for instance, that a byteswap macro, because it is clever, is copyrightable. it's not: whole works are copyrightable. further, the license for the work in question explicitly states that the headers constitute the interface at which the license stops.
there is no issue here. lawyer is trying to make business for himself and others of his species.
BGR leans distinctly towards asininity. in this case, you can omit any mention of android and get a semantically equivalent article. heck, they even quoted the primary source as saying this!
flash: cheap hardware fails more than expensive hardware. wow.
followup: how much of the effect is because owners try harder to avoid damaging expensive hardware?
The ONLY thing Apple has ever done is push the trend towards good graphics. They didn't invent anything, just showed that it could be done well, and that people liked it. the Mac did this, producing a mass-market GUI with reasonably consistent UI rules. the Next basically pushed the resolution and depth of the display, demonstrating the advantage of both. the iPod/iPhone showed that even small displays could use the same basic metaphors with touch.
None of these took place in a vacuum; all of them were extrapolations of work others had done. part of Job's big sell is to convince Appleheads that they were the chosen people, that they had just just a superior product, but a product in a unique category.
Of course Jobs wanted to kill Android - its existence violates the ridiculous marketing mystique he spent billions to create. It's a religious war.
It's also totally immoral. There's simply no way to defend one company saying "no, you must not create good products". And since nothing Apple created came from nowhere, there is no legal basis for claiming some kind of IP monopoly (patent, copyright, trademark, designmark).
Jobs was the Pope of the Church of Apple, and he must have been just as frustrated as Catholic popes were during the reformation.
Maybe the era of industrially manufactured content is just over. Whenever I hear musicians talking about what they do, it's basically albums for marketing purposes with the real revenue product being concerts. This certainly paints record companies as relics of the industrial age, fixated on physical products, not willing to move on.
So how do authors (and/or publishers) move on from the physical-era model of book publishing? Suppose buying a book let you into an online community where you could interact with the book, its supporting material, the author, other readers. Participate in work on followups to the book. In other words, transform from an industry focused on single-physical-product transactions to a model of an ongoing, content-related relationship.
even the "dual-core" tegra2 had a companion core. it's hard to say that this extra management core is a real core, since it's not a peer of the others in, for instance, cache-coherency.
still, sure, asymmetric cores are a nice way to take further advantage of extreme variance in load. even after you've downclocked a normal core as far as it can go, a "designed for slow" core is going to dissipate less power. I'm not sure why supporting this kind of asymmetry would be all that hard for the linux kernel, though.
What's odd in this case is there there's so little respect for science and the scientists that do it. and the idea that the government should hire its own scientists is just absurd - scientists need to report to an academic institution. the interview demonstrates that the agency involved (and this Eric May character) has a giant axe to grind - a political agenda.
agenda is corrosive to science.
but why do so many people feel that they're being misled by scientists? is it just that they don't want to believe what science says?
it's also kind of appalling that they still do these transects with some guys in a bush plane: no continual video record, no constant gps track, etc.
an enumeration has got to be the stupidiest, least usable, most 1950's card-deck kind of approach imaginable.
not that other people (say, librarians or linguists) do a better job. but isn't this really the case where people with no computer experience are effectively designing programs? people with computer/applied-math backgrounds could make this work right.
the US constitution had it right: copyrights and patents exist for the purpose of promoting progress. the primary goal isn't to give people a living - particularly not to guarantee profits to some company. we need to rethink the whole legal infrastructure around the concept of IP...
golly, a phone-call icon that has a phone handset on it. a notebook icon that looks like notebook paper. contact information that looks like a head-and-shoulders of a, um, contact.
the SSL industry is a nasty piece of work - typical extort-what-the-market-will-bear flavor of non-equilibrium capitalism.
all DNS should be PK-signed and encrypted, and SSL should just use pubkeys found in DNS. a domain owner should be able to establish their own keys, signed by the domain key (which is in turn signed by their registrar as part of registration.)
I'll be buying at least one, assuming it's not a botch. To qualify as a botch, Lenovo would need to make the same mistakes as all those no-name $100 tabs: no gpu, insensitive touch, bad display, bad battery life. To win, the tablet doesn't need to even pretend to be an ipad, or for that matter a media player or have any cameras. There is a market for a highly portable (small, 8hr battery, fairly light) touchscreen connectivity (wifi) device. To the user, the most salient aspect of a tablet is the display: it needs to be nice looking (decent IPS, AMOLED), with a modern GPU (snappiness is 99% of the feel of the UI), fairly nice to hold (doesn't have to be CNC-milled spacecraft titanium).
We already have touchscreen thermostats, fridges, home alarm systems, conference-room-status displays, POS terminals. why not just use a cheap android tablet instead? Heck, why not use them for menus at (sit-down) restaurants? Or to keep customers happy when they're having their oil changed or hair cut?
Since the dominant component in all tabs is the display, that's what needs to be optimized. My guess is that integrating touch into the active matrix itself is the main win, though just integrating would eliminate a sheet of glass (material cost, assembly cost, thickness and weight). Cameras don't cost anything, nor do accelerometers, etc. All the teardowns show batteries come after the display/touch assembly, then 3g-type interfaces. (wifi and bluetooth are cheap.) And people: quit the flash-size pissing match: you don't need even 8 GB for a fully-functioning surf-pad. There's no reason for a connectivity tablet to have space for multiple movies - it doesn't have to be a PMP!
Lenovo knows these things, and is not trying to prove anything (unlike, oh, say HP).
it's nice to get a glimpse of the underlying totalitarian foundations of China. what a strange situation - sort of an inverse Potemkin village with most of the country participating in the world economy "for show", while a tiny, reclusive core of government holds all the strings. obviously, that core is not aiming the country on a path of western-style individual liberty. engagement with the West is just a practical technique to fill a billion mouths, to let the west provide advanced technology that can't somehow be withdrawn.
is their thought that some day, China will have gotten enough benefit from the West, and so will stop the charade, flip the switch and resume a normal (for Chinese history) top-down, Mao-wears-the-cap-with-the-ruby-button empire?
Apple produces nicely designed cases, and some of their UI design is tasteful and effective. Otherwise, they suck, but since they drink their own koolaid, they think everyone else sucks, AND that everyone else aspires to be Apple.
if you believe you have designed the uberphone or uberpad (or uber-UI - remember the ancient war with MSFT), it is only natural that you feel threatened whenever someone else makes a vaguely rectangular tablet with, say, a front-facing display.
Apple is letting its hired dogs (lawyers) drive the company's public moves; this is a mistake. lawyers have to sue to justify their existence so you keep them in cages, where they can only growl at your neighbor's lawyers-in-cages.
"design patents" are even more stupid than software patents. I don't see why we allow lobbying of any form, since all we get is bought government.
the manufacturing cost for any 10" capacitative tablet is about $300 - if HP had chosen to sell at $150, it would have been a loss-leader. that's still an intelligent strategy, though obviously only as _part_ of a plan.
HP, like other tablet vendors, seems to have seen Apple making money and said "we want some of that". so they produced a more-or-less comparable tablet and blithely assumed that they could sell it at Apple prices. but regardless of Apps, or whether HP's product was technically at par, we all know that Apple customers pay a significant premium (say 20%) just to be members of Club Apple.
a 7" tablet, at 250 or so would have been a smart place to start, preferably with a big-brother 10" model at 350-400. and yes, being able to run other software stacks would have been another good move - at the very least it would have generated a lot more buzz.
an ad campaign is no substitute for careful and/or intelligent pricing. most pad vendors are just being pricing idiots - that was clearly HP's problem.
the manufacturing cost of any 10", capacitative foo-pad is about $300. (these numbers come from teardowns from companies like isupply, who definitely know what they're talking about.) apple does a good job controlling their supply chain, but they have no magic - any large vendor can do just as well.
so how to sell pads? do something different. make it an 8", instead - the playbook is smart this way, since it winds up with a build cost of closer to $200. since the vast majority of *pad cost is display/touchscreen, here's an idea: integrate them. companies like samsung that already produce LCD panels should be able to integrate touch sensing right into the matrix. saving $40 in this context means being able to hit a retail price $100 lower.
the touchpad fiasco shows us two things. first, $99 is a steal for a more-or-less ipad-equivalent. what's the right price? something more like 300?
of course the other thing HP showed us is what completely inept idiots run the company.
it's a bit hard to understand what the point of this research is. if you actually want to understand neural behavior, simulations are obviously a better path: arbitrarily scalable and more flexible (in reinforcement schedules, etc). if the hope is to produce something more efficient than simulation, great, but where's the stats on fan-in, propagation delay, wire counts, joules-per-op, etc. personally, I find that some people simply have a compulsion to try to replicate neurons in silico - not for any reason, but just because it's their "thing".
worse is the media coverage that loves the very misleading analogy of neurons:transistors. they're actually very dissimilar, and the constraints each operates under are very different.
this is an even better reason we need secure NFC transactions (with your mobile) asap. it's absurd to be typing a by-definition-weak password into an unauditable terminal. why hasn't some bank hasn't noticed that at least early adopters would pay for the privilege of paying securely?
then again, if banks simply secured their terminals, much of the hacked-ATM problem would disappear. yes, toilet-like stalls for each ATM...
tech is just adding power to existing creative tools. auteurs can still produce ad-free art if they wish, and take the risk inherent in getting people to pay for it. more commercial products can customize the product for the watcher, and thus offer content at lower prices.
consumers, for their part, can be as passive as they want, or drive development of software that preprocesses content to remove what they object to (turn all those coke cans to pepsi or guiness).
where's the problem? yes, it means that watching a movie in the future might be quite compute-intensive, but so?
the most interesting consequence is that such tech will eliminate the concept of a finished, static creative work. everything's interactive, malleable, customizable. how does copyright deal with that? the current situation with licensing for sampling music clips is not a viable way forward...
when one of their machines reboots, where does the key come from? such sites usually spend as much money as possible on the theory that mauve is better, which in this case probably means FC SANs. but at which level does the encryption happen? and doesn't disk encryption just mean that you need to take the enclosure or client box too?
Slashdot Mirror
there are lots of useful computations that are more flops-intensive (relative to memory footprint) than dot-products. matmul, fft, almost anything montecarlo, etc.
this "issue" seemed to be mostly a traffic-getting vehicle for phoronix.
this lawyer doesn't really understand programming - even his understanding of copyright is only enough to make him dangerous. he states, for instance, that a byteswap macro, because it is clever, is copyrightable. it's not: whole works are copyrightable. further, the license for the work in question explicitly states that the headers constitute the interface at which the license stops.
there is no issue here. lawyer is trying to make business for himself and others of his species.
short and sweet. of course, all the previous suggestions to name it "defunct" would be even better. for a more pointed suggestion:
http://en.wikipedia.org/wiki/Ministry_of_Public_Enlightenment_and_Propaganda
BGR leans distinctly towards asininity. in this case, you can omit any mention of android and get a semantically equivalent article. heck, they even quoted the primary source as saying this!
flash: cheap hardware fails more than expensive hardware. wow.
followup: how much of the effect is because owners try harder to avoid damaging expensive hardware?
The ONLY thing Apple has ever done is push the trend towards good graphics. They didn't invent anything, just showed that it could be done well, and that people liked it. the Mac did this, producing a mass-market GUI with reasonably consistent UI rules. the Next basically pushed the resolution and depth of the display, demonstrating the advantage of both. the iPod/iPhone showed that even small displays could use the same basic metaphors with touch.
None of these took place in a vacuum; all of them were extrapolations of work others had done. part of Job's big sell is to convince Appleheads that they were the chosen people, that they had just just a superior product, but a product in a unique category.
Of course Jobs wanted to kill Android - its existence violates the ridiculous marketing mystique he spent billions to create. It's a religious war.
It's also totally immoral. There's simply no way to defend one company saying "no, you must not create good products". And since nothing Apple created came from nowhere, there is no legal basis for claiming some kind of IP monopoly (patent, copyright, trademark, designmark).
Jobs was the Pope of the Church of Apple, and he must have been just as frustrated as Catholic popes were during the reformation.
Microsoft spends O(billions) on advertising. $44M on a single product that needs help with traction sounds kind of light, actually.
Maybe the era of industrially manufactured content is just over. Whenever I hear musicians talking about what they do, it's basically albums for marketing purposes with the real revenue product being concerts. This certainly paints record companies as relics of the industrial age, fixated on physical products, not willing to move on.
So how do authors (and/or publishers) move on from the physical-era model of book publishing? Suppose buying a book let you into an online community where you could interact with the book, its supporting material, the author, other readers. Participate in work on followups to the book. In other words, transform from an industry focused on single-physical-product transactions to a model of an ongoing, content-related relationship.
even the "dual-core" tegra2 had a companion core. it's hard to say that this extra management core is a real core, since it's not a peer of the others in, for instance, cache-coherency.
still, sure, asymmetric cores are a nice way to take further advantage of extreme variance in load. even after you've downclocked a normal core as far as it can go, a "designed for slow" core is going to dissipate less power. I'm not sure why supporting this kind of asymmetry would be all that hard for the linux kernel, though.
What's odd in this case is there there's so little respect for science and the scientists that do it. and the idea that the government should hire its own scientists is just absurd - scientists need to report to an academic institution. the interview demonstrates that the agency involved (and this Eric May character) has a giant axe to grind - a political agenda.
agenda is corrosive to science.
but why do so many people feel that they're being misled by scientists? is it just that they don't want to believe what science says?
it's also kind of appalling that they still do these transects with some guys in a bush plane: no continual video record, no constant gps track, etc.
will future history remember Goggle's main significance as hastening the demise of the current patent system?
an enumeration has got to be the stupidiest, least usable, most 1950's card-deck kind of approach imaginable.
not that other people (say, librarians or linguists) do a better job. but isn't this really the case where people with no computer experience are effectively designing programs? people with computer/applied-math backgrounds could make this work right.
the US constitution had it right: copyrights and patents exist for the purpose of promoting progress. the primary goal isn't to give people a living - particularly not to guarantee profits to some company. we need to rethink the whole legal infrastructure around the concept of IP...
golly, a phone-call icon that has a phone handset on it. a notebook icon that looks like notebook paper. contact information that looks like a head-and-shoulders of a, um, contact.
apple sucks. they do evil.
the SSL industry is a nasty piece of work - typical extort-what-the-market-will-bear flavor of non-equilibrium capitalism.
all DNS should be PK-signed and encrypted, and SSL should just use pubkeys found in DNS. a domain owner should be able to establish their own keys, signed by the domain key (which is in turn signed by their registrar as part of registration.)
I'll be buying at least one, assuming it's not a botch. To qualify as a botch, Lenovo would need to make the same mistakes as all those no-name $100 tabs: no gpu, insensitive touch, bad display, bad battery life. To win, the tablet doesn't need to even pretend to be an ipad, or for that matter a media player or have any cameras. There is a market for a highly portable (small, 8hr battery, fairly light) touchscreen connectivity (wifi) device. To the user, the most salient aspect of a tablet is the display: it needs to be nice looking (decent IPS, AMOLED), with a modern GPU (snappiness is 99% of the feel of the UI), fairly nice to hold (doesn't have to be CNC-milled spacecraft titanium).
We already have touchscreen thermostats, fridges, home alarm systems, conference-room-status displays, POS terminals. why not just use a cheap android tablet instead? Heck, why not use them for menus at (sit-down) restaurants? Or to keep customers happy when they're having their oil changed or hair cut?
Since the dominant component in all tabs is the display, that's what needs to be optimized. My guess is that integrating touch into the active matrix itself is the main win, though just integrating would eliminate a sheet of glass (material cost, assembly cost, thickness and weight). Cameras don't cost anything, nor do accelerometers, etc. All the teardowns show batteries come after the display/touch assembly, then 3g-type interfaces. (wifi and bluetooth are cheap.) And people: quit the flash-size pissing match: you don't need even 8 GB for a fully-functioning surf-pad. There's no reason for a connectivity tablet to have space for multiple movies - it doesn't have to be a PMP!
Lenovo knows these things, and is not trying to prove anything (unlike, oh, say HP).
it's nice to get a glimpse of the underlying totalitarian foundations of China. what a strange situation - sort of an inverse Potemkin village with most of the country participating in the world economy "for show", while a tiny, reclusive core of government holds all the strings. obviously, that core is not aiming the country on a path of western-style individual liberty. engagement with the West is just a practical technique to fill a billion mouths, to let the west provide advanced technology that can't somehow be withdrawn.
is their thought that some day, China will have gotten enough benefit from the West, and so will stop the charade, flip the switch and resume a normal (for Chinese history) top-down, Mao-wears-the-cap-with-the-ruby-button empire?
Apple produces nicely designed cases, and some of their UI design is tasteful and effective. Otherwise, they suck, but since they drink their own koolaid, they think everyone else sucks, AND that everyone else aspires to be Apple.
if you believe you have designed the uberphone or uberpad (or uber-UI - remember the ancient war with MSFT), it is only natural that you feel threatened whenever someone else makes a vaguely rectangular tablet with, say, a front-facing display.
Apple is letting its hired dogs (lawyers) drive the company's public moves; this is a mistake. lawyers have to sue to justify their existence so you keep them in cages, where they can only growl at your neighbor's lawyers-in-cages.
"design patents" are even more stupid than software patents. I don't see why we allow lobbying of any form, since all we get is bought government.
the manufacturing cost for any 10" capacitative tablet is about $300 - if HP had chosen to sell at $150, it would have been a loss-leader. that's still an intelligent strategy, though obviously only as _part_ of a plan.
HP, like other tablet vendors, seems to have seen Apple making money and said "we want some of that". so they produced a more-or-less comparable tablet and blithely assumed that they could sell it at Apple prices. but regardless of Apps, or whether HP's product was technically at par, we all know that Apple customers pay a significant premium (say 20%) just to be members of Club Apple.
a 7" tablet, at 250 or so would have been a smart place to start, preferably with a big-brother 10" model at 350-400. and yes, being able to run other software stacks would have been another good move - at the very least it would have generated a lot more buzz.
an ad campaign is no substitute for careful and/or intelligent pricing. most pad vendors are just being pricing idiots - that was clearly HP's problem.
the manufacturing cost of any 10", capacitative foo-pad is about $300. (these numbers come from teardowns from companies like isupply, who definitely know what they're talking about.) apple does a good job controlling their supply chain, but they have no magic - any large vendor can do just as well.
so how to sell pads? do something different. make it an 8", instead - the playbook is smart this way, since it winds up with a build cost of closer to $200. since the vast majority of *pad cost is display/touchscreen, here's an idea: integrate them. companies like samsung that already produce LCD panels should be able to integrate touch sensing right into the matrix. saving $40 in this context means being able to hit a retail price $100 lower.
the touchpad fiasco shows us two things. first, $99 is a steal for a more-or-less ipad-equivalent. what's the right price? something more like 300?
of course the other thing HP showed us is what completely inept idiots run the company.
it's a bit hard to understand what the point of this research is. if you actually want to understand neural behavior, simulations are obviously a better path: arbitrarily scalable and more flexible (in reinforcement schedules, etc). if the hope is to produce something more efficient than simulation, great, but where's the stats on fan-in, propagation delay, wire counts, joules-per-op, etc. personally, I find that some people simply have a compulsion to try to replicate neurons in silico - not for any reason, but just because it's their "thing".
worse is the media coverage that loves the very misleading analogy of neurons:transistors. they're actually very dissimilar, and the constraints each operates under are very different.
this is an even better reason we need secure NFC transactions (with your mobile) asap. it's absurd to be typing a by-definition-weak password into an unauditable terminal. why hasn't some bank hasn't noticed that at least early adopters would pay for the privilege of paying securely?
then again, if banks simply secured their terminals, much of the hacked-ATM problem would disappear. yes, toilet-like stalls for each ATM...
tech is just adding power to existing creative tools. auteurs can still produce ad-free art if they wish, and take the risk inherent in getting people to pay for it. more commercial products can customize the product for the watcher, and thus offer content at lower prices.
consumers, for their part, can be as passive as they want, or drive development of software that preprocesses content to remove what they object to (turn all those coke cans to pepsi or guiness).
where's the problem? yes, it means that watching a movie in the future might be quite compute-intensive, but so?
the most interesting consequence is that such tech will eliminate the concept of a finished, static creative work. everything's interactive, malleable, customizable. how does copyright deal with that? the current situation with licensing for sampling music clips is not a viable way forward...
when one of their machines reboots, where does the key come from? such sites usually spend as much money as possible on the theory that mauve is better, which in this case probably means FC SANs. but at which level does the encryption happen? and doesn't disk encryption just mean that you need to take the enclosure or client box too?