Before the update, Apple did not have voice routing on the iPhone. Now it does. Most people seem to be finding it reasonably reliable for routing to an address, although I've occasionally encountered misrouting on every GPS system I've ever used.
But there are numerous alternatives, both paid and free--including Google Maps, readily accessible from Safari (you can add an icon to your home screen for convenience), which provides the same public transportation routing that the app version previously supplied.
So it is hardly surprising that iPhone owners are not reluctant to update.
Lack of adjustment for multiple comparisons can be a huge problem. If you compare too many things, there is a high likelihood of finding some that meet the "standard" criterion for "statistical significance" purely by chance. There does not seem to be adequate justification for the nonstandard statistical approach taken, and one can't help wondering if they chose this analysis because standard methods, with proper adjustment for multiple comparisons, did not find "statistical significance." Minimally, the issue of multiple comparisons should have been addressed in the discussion, and standard statistical analyses (e.g. ANOVA) should have been provided in addition to the complex regressions.
Another troubling thing is that I could not find any statement that the experiments were done "blind" such that the investigators were unaware of which rats got which feed until after the experiments and analysis were complete. It is very easy to bias your results if you know which was which, even unintentionally. The absence of a clear dose-repsonse relationship is also more suggestive of artifact than a genuine toxicological effect.
I went and upgraded, although I'll miss the transit route-finding. Turn-by-turn is a nice feature, but I bought a third party app for that long ago, so it's not a major selling point for me. But there are plenty of other apps that will tell me when the next bus or train is due, so I'll only miss the transit routes if I have to go somewhere new. And Apple is encouraging other app developers to fill this gap (perhaps even Google, since they were already doing it), so this should be a temporary issue. And there's actually room for improvement on Google's approach. The problem with Google's method is that it would sometimes give you a route that would leave you stranded for an extended period of time if your train or bus ran late and you missed Google's projected connection. I'd love to have an app that could tell me the most robust transit route--the one that would get me to my destination earliest if I missed the projected connection(s). So perhaps this will serve a stimulus to other developers to step into the gap.
Actually, a huge amount of effort has been devoted to curing and preventing cancer, one major cause of chronic pain. So far, we've come up with a couple of quite effective ways to prevent cancer: stop smoking and stay out of the sun. And there are a few cancers, such as Hodgkin's Disease, that can often be cured, and others that can sometimes be cured. But it is a very hard problem--every type of cancer is different, and even in one person, there may be different types of cancer cells, which do not all respond to the same treatment, so it is very hard to eradicate permanently from the body.
That is exactly what is and should be done. In this context, physical dependence is just a medical issue that needs to be properly managed, just as diabetics can be said to be physically dependent on insulin. For the most part, people who take opioids for treatment of chronic pain rather than to "get high" do just fine. They don't tend to escalate their dose excessively, and if their pain issue resolves, they can be tapered off the opiate and are unlikely to seek it out on the street.
Acupuncture may work by creating lots of little cuts/punctures. Each cut is too minor to bother you, but your body still wants to fix them. Each cut is followed by the body's healing responses such as ignoring the pain (I don't know all the terms of the related chemicals). All the healing effects add up and eventually you don't notice your original problem because you're too desensitized. Acupuncture is known to work by the WHO and the NIH for certain types of pain and nausea.
Probably not. There have been controlled studies with sham acupuncture, which uses fake needles that don't penetrate the skin. It works about as well as real acupuncture.
It looks like acupuncture is essentially an elaborate (and generally rather costly) placebo.
Oxycodone, like heroin (which is largely transformed by the body into morphine), morphine, and codeine, acts through mu opiate receptors. All of them do essentially the same thing. All of the drugs of this class are excellent pain relievers (although heroin is not much used for that purpose since it has a bad reputation as a widely abused opioid) and is also highly habit forming, particularly when used recreationally rather than for pain relief.
Time release oxycodone is an important and useful drug for people with severe chronic pain, as it produces pain relief without the user needing to be constantly popping pills.
We have patent laws to reward people and firms for investing time and money in developing new products, and also to reward them for making their discoveries public rather than trying to retain them as trade secrets. It is hard to see what makes software different. You can certainly get a patent for inventing a product that is assembled or built using standard components, which anybody could use to create a similar product (once they have the idea). Like hardware, software may be novel or not, obvious or not. Like hardware, software can add substantial value to an endeavor.
So what, specifically, makes software patents different from hardware patents such that inventors should not be rewarded for their efforts?
Or is it just that you don't like patents in general, and you see innovators in software as more vulnerable to attack?
I paid for Navigon, which is pricey (although cheaper than when I bought it) but as good as the stand-alone units I've used. But you definitely need a power connection for anything more than a short drive.
When traveling, performance can be improved quite a bit by turning off wifi when not in a wifi zone, or turning off cell if you aren't taking calls. GPS is too power hungry to be used more than briefly. If I'm planning to drive and want turn-by-turn, I make sure to bring an auto charger.
I'm sure that there are some enthusiasts and wealthy people who buy every new model, but so far I've not seen an iPhone model that thrilled me so much that I wanted to take the financial hit to upgrade before my 2-year contract expired. I've got my iPad and my iPhone on a staggered upgrade cycle, so I upgrade only one per year. Generally, I find the year-to-year hardware upgrades seem more evolutionary than revolutionary. But upgrading after 2-years always feels like a big advance. I like the bigger screen, but I can wait another year for it, and I know I'll get more goodies too.
I like the cycle I'm on with my iPhone. I skip the first generation with the big new feature (retina display, big screen), and I pick up the next one where they've got it refined and all of the software supports it.
Apple's batteries are essentially state-of-the-art. The trade-off is charge time vs. size, weight, and computing power. Apple clearly feels they have hit the "sweet spot." Some heavy users might feel differently.
Not every display is a list. Apple's solution gives you a pixel-perfect display for older apps, insuring that no apps "break" on the iPhone 5. There is no automated routine that could promise that. And if you've ever looked at Apple's development environment, you know that telling an app to include a longer list on the iPhone 5 will be literally a few minutes worth of work. The only ones where it will take any appreciable effort on the part of developers are the very ones for which an automated routine wouldn't work.
The iPhone 4 does not introduce a new screen resolution--it's exactly the same as the iPhone 4. So a "universal" app (that runs on all iOS devices) needs to accommodate 4 resolutions (same as before) and 3 screen layouts (one more than before). Wow, what terrible fragmentation!
In actual practice, the "fix" turned out to be not clutching your phone too tightly when you are in a marginal signal area. And this turns out to be good advice for all phones without protruding antennas, not just the iPhone. Most people do this pretty much automatically.
Actually, Apple's TrueType fonts included hand-optimized "hints," and even hand-optimized bitmap fonts for the smallest sizes, to deal with the fact that automated methods do not do as good a job of rasterizing as a human can do.
Once you get to "retina" resolution, the difference between hand-optimization and scaling becomes small. But that is irrelevant to the iPhone 5, which uses the same retina resolution as the iPhone 4. Changing the layout of a screen in an automated way to make optimum use of additional screen real estate is a much harder problem than merely scaling an image, because different apps use the space on the screen in very different ways.
I gotta admit, if I were Google, I'd consider not releasing Google Maps for iOS. Apple made their bed--let them lie in it. Why should Google come along and save them?
Right! "We'll teach you a lesson--we won't compete with you! See how you like that!"
Remember, Google doesn't make its money from selling devices like Apple does; it is an advertising company. It makes money whether its apps are on iOS or Android, so long as its customers use them.
If Google bows out, it just makes opportunities for other developers--there are already a great many navigation apps available for iOS.
Yes, there is a reason why most people prefer to buy a car pre-assembled, rather than a box of parts like a LEGO kit. But there will always be hobbyists who would rather do it themselves. That is clearly not the market that Apple's products--especially portable ones like phones, tablets, and laptops--are designed for.
I didn't bother to mention pixel doubling because it doesn't work "just fine." Yes, it's better than trying to scale by a non-integer factor, but it still looks pretty chunky, and was only a stopgap when Retina Displays were introduced. iOS Apps that do not include RD resolution icons are no longer considered acceptable by Apple.
And automatic scaling of non-bitmapped "object graphics" does not work "just fine," either. That's why fonts use "hinting" to optimize display on raster displays. The problem is less on very high resolution displays, so automated scaling may be acceptable on a RD display, but not at anything less than that.
In any case, all of this is quite irrelevant to the iPhone 5, since none of the screen elements need to be scaled, as the resolution is identical to the iPhone 4. The issue is that the additional screen area means that display layout should be adjusted to make optimum use of the additional area--for example, the icon screen now shows an additional row of icons. The best way to do this is very dependent upon the nature of an app's display, so it doesn't make sense to try to do it in an general, automated way. And for almost all apps, it would a trivial job for a human to include a layout for one additional screen size--far less work (and likely a better outcome) than trying a general algorithm to "fit" that app to any conceivable screen shape.
So letterboxing as an interim solution for the few weeks until app developers can post an updated version makes a great deal of sense.
Apple basically does the same thing, but with a more limited range of pixel densities, each one can be exacy tuned to the device. And since the iPhone 5 does not introduce a new pixel density, it is not a factor here
If things just scaled with resolution, then targets would be too small for easy selection on a retina display. Of course, everything can be made resolution-independent so that objects rescale to a constant apparent size, but that will never look quite as nice as a display hand-optimized for a particular resolution.
Yes, it makes it hard for a web designer. Do you design your web page to fill the width of the window dynamically (in which case it is hard to consistently maintain nice composition) or do you pick a preferred width (in which case you probably don't want it to be larger that 640 pixels, which is pretty small on a modern display, for backwards compatibility) and just leave empty space if the user drags the window wider? Like many websites, Slashdot seems to follow a mixed approach: most elements expand to fill the space available, but text entry boxes only expand so far, and if you make shrink the window beyond a certain point, things start to crop. But there's a lot of space on modern computer screens, and users can adjust their window width until things look nice, so it's not such a big deal.
Slashdot Mirror
Before the update, Apple did not have voice routing on the iPhone. Now it does. Most people seem to be finding it reasonably reliable for routing to an address, although I've occasionally encountered misrouting on every GPS system I've ever used.
But there are numerous alternatives, both paid and free--including Google Maps, readily accessible from Safari (you can add an icon to your home screen for convenience), which provides the same public transportation routing that the app version previously supplied.
So it is hardly surprising that iPhone owners are not reluctant to update.
In Safari, go to www.maps.google.com.
Click the middle icon at the bottom of the screen
Select "Add to home screen"
Lack of adjustment for multiple comparisons can be a huge problem. If you compare too many things, there is a high likelihood of finding some that meet the "standard" criterion for "statistical significance" purely by chance. There does not seem to be adequate justification for the nonstandard statistical approach taken, and one can't help wondering if they chose this analysis because standard methods, with proper adjustment for multiple comparisons, did not find "statistical significance." Minimally, the issue of multiple comparisons should have been addressed in the discussion, and standard statistical analyses (e.g. ANOVA) should have been provided in addition to the complex regressions.
Another troubling thing is that I could not find any statement that the experiments were done "blind" such that the investigators were unaware of which rats got which feed until after the experiments and analysis were complete. It is very easy to bias your results if you know which was which, even unintentionally. The absence of a clear dose-repsonse relationship is also more suggestive of artifact than a genuine toxicological effect.
I went and upgraded, although I'll miss the transit route-finding. Turn-by-turn is a nice feature, but I bought a third party app for that long ago, so it's not a major selling point for me. But there are plenty of other apps that will tell me when the next bus or train is due, so I'll only miss the transit routes if I have to go somewhere new. And Apple is encouraging other app developers to fill this gap (perhaps even Google, since they were already doing it), so this should be a temporary issue. And there's actually room for improvement on Google's approach. The problem with Google's method is that it would sometimes give you a route that would leave you stranded for an extended period of time if your train or bus ran late and you missed Google's projected connection. I'd love to have an app that could tell me the most robust transit route--the one that would get me to my destination earliest if I missed the projected connection(s). So perhaps this will serve a stimulus to other developers to step into the gap.
Actually, a huge amount of effort has been devoted to curing and preventing cancer, one major cause of chronic pain. So far, we've come up with a couple of quite effective ways to prevent cancer: stop smoking and stay out of the sun. And there are a few cancers, such as Hodgkin's Disease, that can often be cured, and others that can sometimes be cured. But it is a very hard problem--every type of cancer is different, and even in one person, there may be different types of cancer cells, which do not all respond to the same treatment, so it is very hard to eradicate permanently from the body.
That is exactly what is and should be done. In this context, physical dependence is just a medical issue that needs to be properly managed, just as diabetics can be said to be physically dependent on insulin. For the most part, people who take opioids for treatment of chronic pain rather than to "get high" do just fine. They don't tend to escalate their dose excessively, and if their pain issue resolves, they can be tapered off the opiate and are unlikely to seek it out on the street.
Coming up with a good time release formulation can be quite tricky. It is certainly reasonable that it can be patented.
Probably not. There have been controlled studies with sham acupuncture, which uses fake needles that don't penetrate the skin. It works about as well as real acupuncture.
It looks like acupuncture is essentially an elaborate (and generally rather costly) placebo.
Oxycodone, like heroin (which is largely transformed by the body into morphine), morphine, and codeine, acts through mu opiate receptors. All of them do essentially the same thing. All of the drugs of this class are excellent pain relievers (although heroin is not much used for that purpose since it has a bad reputation as a widely abused opioid) and is also highly habit forming, particularly when used recreationally rather than for pain relief.
Time release oxycodone is an important and useful drug for people with severe chronic pain, as it produces pain relief without the user needing to be constantly popping pills.
We have patent laws to reward people and firms for investing time and money in developing new products, and also to reward them for making their discoveries public rather than trying to retain them as trade secrets. It is hard to see what makes software different. You can certainly get a patent for inventing a product that is assembled or built using standard components, which anybody could use to create a similar product (once they have the idea). Like hardware, software may be novel or not, obvious or not. Like hardware, software can add substantial value to an endeavor.
So what, specifically, makes software patents different from hardware patents such that inventors should not be rewarded for their efforts?
Or is it just that you don't like patents in general, and you see innovators in software as more vulnerable to attack?
Good that Apple has decided to go with a connector that preserves much of the functionality of the old connector, while providing the smaller form factor required by the new devices (with a micro-USB adaptor available for the European market, which ensures compliance with the standard).
I paid for Navigon, which is pricey (although cheaper than when I bought it) but as good as the stand-alone units I've used. But you definitely need a power connection for anything more than a short drive.
When traveling, performance can be improved quite a bit by turning off wifi when not in a wifi zone, or turning off cell if you aren't taking calls. GPS is too power hungry to be used more than briefly. If I'm planning to drive and want turn-by-turn, I make sure to bring an auto charger.
I'm sure that there are some enthusiasts and wealthy people who buy every new model, but so far I've not seen an iPhone model that thrilled me so much that I wanted to take the financial hit to upgrade before my 2-year contract expired. I've got my iPad and my iPhone on a staggered upgrade cycle, so I upgrade only one per year. Generally, I find the year-to-year hardware upgrades seem more evolutionary than revolutionary. But upgrading after 2-years always feels like a big advance. I like the bigger screen, but I can wait another year for it, and I know I'll get more goodies too.
I like the cycle I'm on with my iPhone. I skip the first generation with the big new feature (retina display, big screen), and I pick up the next one where they've got it refined and all of the software supports it.
Apple's batteries are essentially state-of-the-art. The trade-off is charge time vs. size, weight, and computing power. Apple clearly feels they have hit the "sweet spot." Some heavy users might feel differently.
Not on my AT&T 4s
Not every display is a list. Apple's solution gives you a pixel-perfect display for older apps, insuring that no apps "break" on the iPhone 5. There is no automated routine that could promise that. And if you've ever looked at Apple's development environment, you know that telling an app to include a longer list on the iPhone 5 will be literally a few minutes worth of work. The only ones where it will take any appreciable effort on the part of developers are the very ones for which an automated routine wouldn't work.
The iPhone 4 does not introduce a new screen resolution--it's exactly the same as the iPhone 4. So a "universal" app (that runs on all iOS devices) needs to accommodate 4 resolutions (same as before) and 3 screen layouts (one more than before). Wow, what terrible fragmentation!
In actual practice, the "fix" turned out to be not clutching your phone too tightly when you are in a marginal signal area. And this turns out to be good advice for all phones without protruding antennas, not just the iPhone. Most people do this pretty much automatically.
Actually, Apple's TrueType fonts included hand-optimized "hints," and even hand-optimized bitmap fonts for the smallest sizes, to deal with the fact that automated methods do not do as good a job of rasterizing as a human can do.
Once you get to "retina" resolution, the difference between hand-optimization and scaling becomes small. But that is irrelevant to the iPhone 5, which uses the same retina resolution as the iPhone 4. Changing the layout of a screen in an automated way to make optimum use of additional screen real estate is a much harder problem than merely scaling an image, because different apps use the space on the screen in very different ways.
Right! "We'll teach you a lesson--we won't compete with you! See how you like that!"
Remember, Google doesn't make its money from selling devices like Apple does; it is an advertising company. It makes money whether its apps are on iOS or Android, so long as its customers use them.
If Google bows out, it just makes opportunities for other developers--there are already a great many navigation apps available for iOS.
Yes, there is a reason why most people prefer to buy a car pre-assembled, rather than a box of parts like a LEGO kit. But there will always be hobbyists who would rather do it themselves. That is clearly not the market that Apple's products--especially portable ones like phones, tablets, and laptops--are designed for.
I didn't bother to mention pixel doubling because it doesn't work "just fine." Yes, it's better than trying to scale by a non-integer factor, but it still looks pretty chunky, and was only a stopgap when Retina Displays were introduced. iOS Apps that do not include RD resolution icons are no longer considered acceptable by Apple.
And automatic scaling of non-bitmapped "object graphics" does not work "just fine," either. That's why fonts use "hinting" to optimize display on raster displays. The problem is less on very high resolution displays, so automated scaling may be acceptable on a RD display, but not at anything less than that.
In any case, all of this is quite irrelevant to the iPhone 5, since none of the screen elements need to be scaled, as the resolution is identical to the iPhone 4. The issue is that the additional screen area means that display layout should be adjusted to make optimum use of the additional area--for example, the icon screen now shows an additional row of icons. The best way to do this is very dependent upon the nature of an app's display, so it doesn't make sense to try to do it in an general, automated way. And for almost all apps, it would a trivial job for a human to include a layout for one additional screen size--far less work (and likely a better outcome) than trying a general algorithm to "fit" that app to any conceivable screen shape.
So letterboxing as an interim solution for the few weeks until app developers can post an updated version makes a great deal of sense.
Apple basically does the same thing, but with a more limited range of pixel densities, each one can be exacy tuned to the device. And since the iPhone 5 does not introduce a new pixel density, it is not a factor here
If things just scaled with resolution, then targets would be too small for easy selection on a retina display. Of course, everything can be made resolution-independent so that objects rescale to a constant apparent size, but that will never look quite as nice as a display hand-optimized for a particular resolution.
Yes, it makes it hard for a web designer. Do you design your web page to fill the width of the window dynamically (in which case it is hard to consistently maintain nice composition) or do you pick a preferred width (in which case you probably don't want it to be larger that 640 pixels, which is pretty small on a modern display, for backwards compatibility) and just leave empty space if the user drags the window wider? Like many websites, Slashdot seems to follow a mixed approach: most elements expand to fill the space available, but text entry boxes only expand so far, and if you make shrink the window beyond a certain point, things start to crop. But there's a lot of space on modern computer screens, and users can adjust their window width until things look nice, so it's not such a big deal.