True to some degree. But many, even most, of those trips (go charge somewhere, transports) would happen with or without self-driving cars. And it's not clear to me that a car going directly to a somewhat distant parking garage is actually using any more energy than a car circling for fifteen minutes looking for a convenient nearby spot.
The smoother, slower ride of an unhurried automatic car will use less energy than the jerky stop-and-go of impatient human drivers. And as they act as pace cars there'll be a positive effect on all drivers, not just on the automatic cars.
So while your point is something to be concerned about I suspect the actual impact would be much less than you'd think at first.
* The car can park itself anywhere, get service or pick up stuff while you're working. Less need to use valuable city real estate and street area on parking. And as people no longer park along the streets they get effectively wider, with more space for traffic but also for bicycle lanes.
* A two-car family may only need one, as the car can go by itself to pick up family members as they need it.
* A family may in fact own no car. Car pooling becomes much more effective when you can call up a car from the pool to your front door at any time.
* No need for a license. People with dementia, or taking medication, or with severe disabilities, or underage can still get around, no problem.
* The cars will be scrupulous about obeying traffic laws and speed limits. But even with a small part self-driving cars, they will act as pace cars and slow and smooth traffic for everyone. Even more so, as they'll be recording everything happening around them, and other drivers know it. Pace will be slower, but people will arrive sooner.
* Life becomes tough for taxi drivers. Taking a taxi would become the same as short-term car rental in practice, and cheaper than taxis as there's no drivers salary to pay.
* Point to point transport becomes cheaper too, with driverless vans and trucks shuttling between shipment centers.
* Driverless drive-ins means you can send a car to do a lot of your errands.
Re:Rattle for R baseed datamining
on
R In a Nutshell
·
· Score: 1
I had no idea about rattle. It looks very useful; thanks for the info.
They sell some titles through the Android app store - but of course, that would probably count as a direct sale too. I'd like to get this book there but for some reason they mostly seem to sell Windows-related titles and not much else.
Don't like Coca-Cola or Pepsi? Gone. Apple's turtlenecky hipness is getting cloying? Never hear of them or their fanboys again.
And of course, if you're one of the less morally scrupulous companies out there (as in, pretty much all of them), it could be very tempting to have a botnet provider silently install a filter that removes any trace of your competition on a few million computers out there.
Oh? Which phones? I wasn't aware that there was a single Symbian phone on sale in Japan; all the manufacturers seem to do their own in-house systems. If any of them actually use Symbian they're certainly not mentioning it publicly.
"Many laws have been repealed and governments overthrown because people were more concerned with doing what was right than doing what was legal."
But the real change - the change that counts in the end - happened when those laws were repealed, not when people were flouting them, unjust as they may be. Slavery didn't end when people ignored or broke laws condoning the practice, but when those laws were finally gone.
The problem is when you have a larger system, with hundreds of cores, and an iterative simulation. You run the system for a cycle, propagate data, then run for another cycle and so on. In that case you can't isolate a long-running process on the card, and you end up having to squeeze data through that bus for each cycle anyway. It is likely still well worth using GPUs but you do need to take a good look at whether adding GPUs are more or less effective than using your funds to simply add more cores instead.
I expect over time to see better suited interfaces appear for this type of computing.
"So to get good performance, the problem needs to fit on the GPU. You can move data to and from the main memory (or disk) occasionally, but most of the crunching must happen on card."
From what I have seen when people use GPUs for HPC, this, more often than anything else, is the limiting factor. The actual calculations are plenty fast, but the need to format your data for the GPU, send it, then do the same in reverse for the result really limits the practical gain you get.
I'm not saying it's useless or anything - far from it - but this issue is as important asthe actual processing you want to do for determining what kind of gain you'll see from such an approach.
"While I totally share the view that Ebooks are much less worth to the buyer (me) than "real" books, they're hardly cheaper to produce than paper books. "
I don't doubt you for a second.
But as a customer the reality is still that the ebook gives me a lot less value than the printed version. There is a solution, however. The marginal cost of distributing the ebook is just about zero once you've created it. So sell them as a bundle. Add, oh, 10% to the price of the paper book, and give me access to the ebook version as well.
Yes, I've tried shapewriter; that's the basis of comparison for me. Swedish and English is available, but no Japanese. Anyway, that's not the main reason I don't use it. I just prefer flick input myself. Different strokes - literally:).
"They are faster than multi-tap for sure, but, with the exception of typing obscure non-dictionary words, swype will be way faster."
In my particular case (admittedly not very common) I write in multiple languages. An input method that relies on dictionary matching would require me to switch the input language all the time. You could solve it by allowing the use of multiple simultaneous dictionaries of course, but I know of no input system that would allow me to do so (and there may be reliability problems if you increase the corpus too much I guess).
But I also like flick better because it lets me type with the same hand I hold the phone. Swype - and, to a lesser extent, a normal qwerty input - really needs you to hold the phone with one hand and type with the other.
Anyway, I really like that there's multiple ways of typing available - not to mention Android phones with hardware numpads, slideout keyboards and laptop-style fold-out designs. We can all choose our own preferred way to write.
Of those two I like OpenWnn better, but both do suffer a bit from featuritis. They have a system of plugins where you can add all kinds of shortcuts, precanned sentences and stuff. The default Japanese Xperia keyboard is much cleaner and simpler, and feels more responsive too.
I really prefer flick input, to be honest - 9-key layout, and a small "drag" in different directions determine which character to enter. You got big, easy to hit keys, and you're not dependent on the predictive input to get words right. I started using it for Japanese, but it's so convenient I now use it for English as well. Now, if I could add the Swedish characters to the Xperia Japanese/English flick keyboard I would never have to switch input method again.
I live here and commute every day. I see iPhones - and now, Android phones - every day. Yes, all smartphones combined are just a smaller part of the overall cellphone market; but then, that is the case in the US and Europe as well, all the marketing hype notwithstanding.
"First of all smart phones (as we in the US know them) are almost non-existant in Japan. "
The iPhone has been here for years and is a major hit. The Sony Ericsson Xperia Android phone is NTT Docomo's best selling phone in recent history. You can also get Windows-based smartphones and Blackberrys, though they're obviously aimed squarely at the suit-and-tie set. And now the iPhone 4 is being heavily preordered while every carrier is coming out with Android models as fast as they can. So no, smsrtphones are not "almost non-existent" in Japan.
With that said, and as much as I love my android phone, you're right: if you want portable gaming, today's smartphones don't really hold a candle to a dedicated gaming device. Beginning with the controls, a dedicated device gives you a better game than a phone, and the addition of 3D is just going to widen that gap. A very common sight on the morning commute is people using their phones as a walkman while they play on a DS2.
That's one thing that could be improved I agree. What I do is edit the image in whatever way I wanted to. Then I go to "Photo->New Version" to make a new copy of the image, right-click on the image and copy the place (the qualified filename) and then, in a terminal simply copy the edited one I made to the new version.
One thing that could help (apart from making a simpler plugin interface) would be to be able to simply tell F-spot that "this image is actually just a version of that one". Could be as easy as drag and drop one image onto another to add a copy of it as a version of the target. The opposite - dragging one image in a stack out between other images - would make a copy as a separate image. That way you could use any tools at all, import the image, then just drop it where it belongs.
There also seems to be no support for hierarchical tags or for having many tags in general, just a linear dump of all tags you've got. Not so much fun when you have tags in the many hundreds, and when you want one tag to actually generate two or more tags in the final taglist.
And little to no support for having multiple versions of an image; the only thing seems to be this: "Shotwell stores your edits in a database and applies them on the fly as necessary.". Which is great fun, I guess, if your original image is a 300Mb MF film-scan and you have to wait for 30 seconds while edits are applied whenever you want to see another of your versions. In fact, doesn't this feature pretty much preclude using external tools altogether?
F-spot is pretty stable for me by now, and it can cope with the amount of images I have in a way that no other organizer tool I've tried on linux can. There's a few missing features still - a "light table" mode, where you can collect and compare a set of images directly would be great - but it works fairly well if you have to keep track of a largish number of images from very different sources.
"It's also a disingenuous way to represent the current culture climate."
Last time I looked, newspapers were into reporting news. "Represent the current culture climate" is what literature majors are supposed to be doing between shifts at the fryer station.
I'm mainly talking about the reaction among faculty, rather than the university. Specifically, there were no organized calls for boycotting all Elseviers publications, refusing to serve as editors or reviewers or ignoring papers in them for the purpose of job search or promotion.
Also, there's a fair amount of Japanese vocabulary that's common but only used in writing (I get an earful from my wife whenever I try to use it speaking with her), and I suspect one reason may be that those words are too easily mixed up or misunderstood without the written clues.
"A Kanji usually only has one meaning. Only the attempt to translate it into a foreign language gives it several meanings."
Um, no. They'd sometimes have only one meaning. They sometimes have several. Also, the same word, the same meaning can at times be represented by more than one character with only very subtle (or no) shade of difference.
The other thing to keep in mind is that when characters are used in compounds, their meaning will sometimes carry through, and sometimes their meaning will have nothing at all to do with the meaning of the compound. They may have been chosen for historical reasons, or (embarrasingly often) their use originated in a miscopying - a spelling error, essentially - that became the norm.
Slashdot Mirror
True to some degree. But many, even most, of those trips (go charge somewhere, transports) would happen with or without self-driving cars. And it's not clear to me that a car going directly to a somewhat distant parking garage is actually using any more energy than a car circling for fifteen minutes looking for a convenient nearby spot.
The smoother, slower ride of an unhurried automatic car will use less energy than the jerky stop-and-go of impatient human drivers. And as they act as pace cars there'll be a positive effect on all drivers, not just on the automatic cars.
So while your point is something to be concerned about I suspect the actual impact would be much less than you'd think at first.
* The car can park itself anywhere, get service or pick up stuff while you're working. Less need to use valuable city real estate and street area on parking. And as people no longer park along the streets they get effectively wider, with more space for traffic but also for bicycle lanes.
* A two-car family may only need one, as the car can go by itself to pick up family members as they need it.
* A family may in fact own no car. Car pooling becomes much more effective when you can call up a car from the pool to your front door at any time.
* No need for a license. People with dementia, or taking medication, or with severe disabilities, or underage can still get around, no problem.
* The cars will be scrupulous about obeying traffic laws and speed limits. But even with a small part self-driving cars, they will act as pace cars and slow and smooth traffic for everyone. Even more so, as they'll be recording everything happening around them, and other drivers know it. Pace will be slower, but people will arrive sooner.
* Life becomes tough for taxi drivers. Taking a taxi would become the same as short-term car rental in practice, and cheaper than taxis as there's no drivers salary to pay.
* Point to point transport becomes cheaper too, with driverless vans and trucks shuttling between shipment centers.
* Driverless drive-ins means you can send a car to do a lot of your errands.
I had no idea about rattle. It looks very useful; thanks for the info.
They sell some titles through the Android app store - but of course, that would probably count as a direct sale too. I'd like to get this book there but for some reason they mostly seem to sell Windows-related titles and not much else.
"The marketing possibilities are endless."
As are the anti-marketing possibilities.
Don't like Coca-Cola or Pepsi? Gone. Apple's turtlenecky hipness is getting cloying? Never hear of them or their fanboys again.
And of course, if you're one of the less morally scrupulous companies out there (as in, pretty much all of them), it could be very tempting to have a botnet provider silently install a filter that removes any trace of your competition on a few million computers out there.
"Symbian^2 is fairly popular in Japan, "
Oh? Which phones? I wasn't aware that there was a single Symbian phone on sale in Japan; all the manufacturers seem to do their own in-house systems. If any of them actually use Symbian they're certainly not mentioning it publicly.
"Many laws have been repealed and governments overthrown because people were more concerned with doing what was right than doing what was legal."
But the real change - the change that counts in the end - happened when those laws were repealed, not when people were flouting them, unjust as they may be. Slavery didn't end when people ignored or broke laws condoning the practice, but when those laws were finally gone.
So yes, it is the legality that counts.
"There is a difference between morality and legality. Learn it."
The difference being, it's the legality that counts.
Kan = Japanese
Can = beer coming right up
The problem is when you have a larger system, with hundreds of cores, and an iterative simulation. You run the system for a cycle, propagate data, then run for another cycle and so on. In that case you can't isolate a long-running process on the card, and you end up having to squeeze data through that bus for each cycle anyway. It is likely still well worth using GPUs but you do need to take a good look at whether adding GPUs are more or less effective than using your funds to simply add more cores instead.
I expect over time to see better suited interfaces appear for this type of computing.
"So to get good performance, the problem needs to fit on the GPU. You can move data to and from the main memory (or disk) occasionally, but most of the crunching must happen on card."
From what I have seen when people use GPUs for HPC, this, more often than anything else, is the limiting factor. The actual calculations are plenty fast, but the need to format your data for the GPU, send it, then do the same in reverse for the result really limits the practical gain you get.
I'm not saying it's useless or anything - far from it - but this issue is as important asthe actual processing you want to do for determining what kind of gain you'll see from such an approach.
"While I totally share the view that Ebooks are much less worth to the buyer (me) than "real" books, they're hardly cheaper to produce than paper books. "
I don't doubt you for a second.
But as a customer the reality is still that the ebook gives me a lot less value than the printed version. There is a solution, however. The marginal cost of distributing the ebook is just about zero once you've created it. So sell them as a bundle. Add, oh, 10% to the price of the paper book, and give me access to the ebook version as well.
Yes, I've tried shapewriter; that's the basis of comparison for me. Swedish and English is available, but no Japanese. Anyway, that's not the main reason I don't use it. I just prefer flick input myself. Different strokes - literally :).
"They are faster than multi-tap for sure, but, with the exception of typing obscure non-dictionary words, swype will be way faster."
In my particular case (admittedly not very common) I write in multiple languages. An input method that relies on dictionary matching would require me to switch the input language all the time. You could solve it by allowing the use of multiple simultaneous dictionaries of course, but I know of no input system that would allow me to do so (and there may be reliability problems if you increase the corpus too much I guess).
But I also like flick better because it lets me type with the same hand I hold the phone. Swype - and, to a lesser extent, a normal qwerty input - really needs you to hold the phone with one hand and type with the other.
Anyway, I really like that there's multiple ways of typing available - not to mention Android phones with hardware numpads, slideout keyboards and laptop-style fold-out designs. We can all choose our own preferred way to write.
The keyboard I use is the one included in the Japanese Xperia, and is actually the best one I've used so far. Simeji (http://www.appbrain.com/app/com.adamrocker.android.input.simeji/) and OpenWnn (http://www.appbrain.com/app/com.owplus.ime.openwnnplus) are similar Japanese input methods, and both have several modes including flick input (the left screenshot in the Simeji link and right screenshot for OpenWnn shows flick input in action for Japanese).
Of those two I like OpenWnn better, but both do suffer a bit from featuritis. They have a system of plugins where you can add all kinds of shortcuts, precanned sentences and stuff. The default Japanese Xperia keyboard is much cleaner and simpler, and feels more responsive too.
I really prefer flick input, to be honest - 9-key layout, and a small "drag" in different directions determine which character to enter. You got big, easy to hit keys, and you're not dependent on the predictive input to get words right. I started using it for Japanese, but it's so convenient I now use it for English as well. Now, if I could add the Swedish characters to the Xperia Japanese/English flick keyboard I would never have to switch input method again.
I live here and commute every day. I see iPhones - and now, Android phones - every day. Yes, all smartphones combined are just a smaller part of the overall cellphone market; but then, that is the case in the US and Europe as well, all the marketing hype notwithstanding.
"First of all smart phones (as we in the US know them) are almost non-existant in Japan. "
The iPhone has been here for years and is a major hit. The Sony Ericsson Xperia Android phone is NTT Docomo's best selling phone in recent history. You can also get Windows-based smartphones and Blackberrys, though they're obviously aimed squarely at the suit-and-tie set. And now the iPhone 4 is being heavily preordered while every carrier is coming out with Android models as fast as they can. So no, smsrtphones are not "almost non-existent" in Japan.
With that said, and as much as I love my android phone, you're right: if you want portable gaming, today's smartphones don't really hold a candle to a dedicated gaming device. Beginning with the controls, a dedicated device gives you a better game than a phone, and the addition of 3D is just going to widen that gap. A very common sight on the morning commute is people using their phones as a walkman while they play on a DS2.
"is he being decoded as well?"
He's already pickled; they can wait and do that at any time.
That's one thing that could be improved I agree. What I do is edit the image in whatever way I wanted to. Then I go to "Photo->New Version" to make a new copy of the image, right-click on the image and copy the place (the qualified filename) and then, in a terminal simply copy the edited one I made to the new version.
One thing that could help (apart from making a simpler plugin interface) would be to be able to simply tell F-spot that "this image is actually just a version of that one". Could be as easy as drag and drop one image onto another to add a copy of it as a version of the target. The opposite - dragging one image in a stack out between other images - would make a copy as a separate image. That way you could use any tools at all, import the image, then just drop it where it belongs.
There also seems to be no support for hierarchical tags or for having many tags in general, just a linear dump of all tags you've got. Not so much fun when you have tags in the many hundreds, and when you want one tag to actually generate two or more tags in the final taglist.
And little to no support for having multiple versions of an image; the only thing seems to be this: "Shotwell stores your edits in a database and applies them on the fly as necessary.". Which is great fun, I guess, if your original image is a 300Mb MF film-scan and you have to wait for 30 seconds while edits are applied whenever you want to see another of your versions. In fact, doesn't this feature pretty much preclude using external tools altogether?
F-spot is pretty stable for me by now, and it can cope with the amount of images I have in a way that no other organizer tool I've tried on linux can. There's a few missing features still - a "light table" mode, where you can collect and compare a set of images directly would be great - but it works fairly well if you have to keep track of a largish number of images from very different sources.
"It's also a disingenuous way to represent the current culture climate."
Last time I looked, newspapers were into reporting news. "Represent the current culture climate" is what literature majors are supposed to be doing between shifts at the fryer station.
I'm mainly talking about the reaction among faculty, rather than the university. Specifically, there were no organized calls for boycotting all Elseviers publications, refusing to serve as editors or reviewers or ignoring papers in them for the purpose of job search or promotion.
Also, there's a fair amount of Japanese vocabulary that's common but only used in writing (I get an earful from my wife whenever I try to use it speaking with her), and I suspect one reason may be that those words are too easily mixed up or misunderstood without the written clues.
"A Kanji usually only has one meaning. Only the attempt to translate it into a foreign language gives it several meanings."
Um, no. They'd sometimes have only one meaning. They sometimes have several. Also, the same word, the same meaning can at times be represented by more than one character with only very subtle (or no) shade of difference.
For example, "dai/shiro" ( http://jlex.org/dictionary/1982860 and http://jlex.org/dictionary/1411560 ) has several meanings that are all but unrelated to each other, ranging from price, era, change, (paper) margin and so on. And it's not at all uncommon.
The other thing to keep in mind is that when characters are used in compounds, their meaning will sometimes carry through, and sometimes their meaning will have nothing at all to do with the meaning of the compound. They may have been chosen for historical reasons, or (embarrasingly often) their use originated in a miscopying - a spelling error, essentially - that became the norm.