Juice had better not have added sugar. If it does, it isn't juice. Don't be fooled by things that don't say that they are juice (e.g. Kern's "Nectar", which is almost entirely high fructose corn syrup).
As for whether juice is bad for you, a lot of that depends on whether it brings dietary fiber along with it. Opt for pressed/unfiltered apple juice rather than the filtered stuff, for example.
For that matter, Lean Cuisine meals are about $2.00 apiece at Target, and qualify as a reasonably healthy meal. Double that at a grocery, but even still, it's not significantly more expensive than a 1/4 pounder, fries, and a Coke.
That's not quite right. The answers to FB quizzes aren't stored as part of your profile. AFAICT, FB developers don't have access to any storage as part of the FB platform. That's one of the things I dislike about the concept of writing FB apps: it's BYOS (bring your own storage).
Thus, if the results are stored at all, they are stored by the application developer. It may be tied to your profile ID, but it is not part of your profile. Only the text that it spits out at the end is stored as part of your profile, and only if you tell it to add that result to your wall or whatever.
I realize I'm being pedantic here and that it makes little difference in practice, but....
The IQ test is not a Facebook app, and people aren't really sending it to you. That's just an ad styled to look like it is coming from Facebook. Real Facebook apps have access to your phone number from your profile and wouldn't need to ask.
No, most of the tests are user-generated things created with one of a couple of "quiz generator" applications. The IQ test, AFAIK, is not a Facebook app, though its ads and stylemake it look like one. It's a scam to get people to give them their cell phone numbers.
Likewise. Then again, I'm caucasian with only medium-dark hair. Call me cynical, but $10 says that if somebody with dark hair, darker skin and Middle-Eastern features did the same thing, there would be plainclothes security people casually walking by but watching that person really carefully within two minutes.
If you get enough sunlight (or artificial equivalent), you can produce all the vitamin D you need yourself. So this basically boils down to "energy" again. In fact, AFAIK, you basically can't get enough vitamin D from your diet alone unless the foods or drinks you consume have been artificially enriched (e.g. milk). Even if you could get enough from your diet, you'd probably end up massively overdosing on other vitamins.:-)
According to one of the sample headers that ships with HeaderDoc, TrainsCanFly.
/*!
@category Trains_2(TrainsCanFly)
@abstract The TrainsCanFly category adds levitation methods to the Train class.
@discussion Methods declared in the TrainsCanFly category of the Train class can only be used with properly equiped Train objects.
*/
@interface Trains_2(TrainsCanFly) /*!
@method levitateToHeight:
@abstract Raises the train specified number of centimeters
@param height The number of centimeters to levitate.
*/
- (void)levitateToHeight:(float)height;
@end
There are actually a fair number of iPhone apps on the store that provide DAV or HTTP servers for downloading files from the phone, so a BitTorrent client isn't exactly infeasible. It wouldn't be particularly useful given the limited amount of storage on the device, of course.... And as others have noted, this is basically just a native UI skin for a web interface to a BitTorrent client on another machine, not an actual client. Still, I'd probably be worried about contributory infringement claims if somebody tried to host such an app on a site I ran, so I can see why Apple would shy away from it.
That said, I'm left wondering why the developers of this application didn't just write it as a web app. If you take advantage of the HTML5 offline application cache to allow it to be cached on disk (flash), there's really not much benefit to writing a native app for something this trivial. That would avoid the whole question of whether Apple would allow it on the app store.
I think I see part of the confusion in this thread. I've read some of the papers and they confused me at first. There are two terms called a "block". There is a logical block (512 bytes) and there is a flash block, which is a group of flash pages that must be erased together. Flash pages can be reordered within flash blocks, at least if you're talking about temporary log blocks. I haven't found any evidence that reordering is done in 512-byte quantities. Technical papers tend to use the word "sector" to refer to a 512-byte block to minimize confusion caused by the massively overloaded term "block".
As an aside, this all points to a need for trained technical writers/editors to be involved in the crafting of technical papers; a trained writer or editor would likely have immediately complained about the reuse of the word "block". Flash is really a painful edge case in that the physical block size is larger than the logical block size, which is just plain wrong conceptually. Someone would have insisted on a more precise name like "page group" instead of "flash block". This confusion should have been nipped in the bud years ago....:-)
BTW, I *think* (but I'm not certain) that the linked post is confusing block storage with log storage. AFAIK, the only truly fragmented flash blocks are out-of-band log blocks that contain short-term copies of flash pages that were recently written. Eventually, those writes get flushed out to their normal places within standard, page-mapped flash pages. It is basically a highly persistent write cache.
The controller doesn't have to work around that. When those get full, the still-pertinent data in older log blocks has to be flushed, so yes, there's a performance hit---it degrades to the true random write performance of the device, give or take---but not because of fragmentation. The performance hit is because the cache is full. AFAIK, the controller should not generally be erasing these and rewriting them with some blocks replaced. That would defeat the whole purpose of using log blocks in the first place, which is to minimize the number of erasures during writes to improve burst write performance.
I guess you might make the argument that erasing and rewriting it would be faster if most of the cells in the log block are no longer valid (which in real-world use would probably mostly involve metadata blocks that get rewritten frequently), and I'd imagine that some controllers do perform that sort of tradeoff analysis, but I would expect that in anything approaching a truly random write test, the controller would be flushing far more often than consolidating unless your OS disk cache is really dumb/nonexistent.:-)
Unless I'm misunderstanding things, of course, which is entirely possible....
Yes, I know that extra NAND storage is traditionally used for the mapping. I was initially thinking this would increase write amplification by an order of magnitude or more until the device was full. I have since thought through this a few dozen more times and convinced myself that this could be avoided.
Come again? I'm not aware of SSDs doing any mapping at that level of granularity; to do so would mean that a 256 GB hard drive would require something like 2 GB of storage (assuming an even 32-bit LBA) just to hold the sector mapping table, and that would have to be NOR flash or single-level NAND flash, making it really freakishly expensive.
All SSDs that I'm aware of work like this: you have a logical block address (sector number). These are grouped (contiguously) into pages (those 128KB "slices" you're referred to). If you have a 128 MB page size, then sectors 0-255 would be logical page 0, sectors 1-511 are logical page 1, and so on.
These logical pages are then arbitrarily mapped to physical pages. This results in a mapping table that is a much more plausible 8 megabytes in size for a 256 GB flash drive.
Thus, any fragmentation within a flash page is caused entirely by the filesystem being fragmented, not by anything happening in the flash controller. The performance degradation from full flash has nothing to do with fragmentation. It is caused by the flash controller having to erase an already-used page before it writes to it. The only way to avoid that is to erase unmapped spare pages ahead of time.
Wow. I ran that on President Bush, and the first paragraph was actually pretty close to comprehensible and on-target....
No other topic is more important and explains better the demise of our society than the saga of Pres. George W Bush. Here's the story: An understanding of the damage that may be caused by Pres. Bush's conniving personal attacks isn't something I expect everyone to develop the first time they hear about it. That's why I write over and over again and from so many different angles about how if we don't reach out to the poor, the marginalized, and those unfortunate enough to have been labeled as scornful by Pres. Bush's propaganda machine, our children will curse us in our graves. Speaking of our children, we need to teach them diligently that as that last sentence suggests, Pres. Bush argues that he should demand that loyalty to asinine backbiters supersedes personal loyalty because "it's the right thing to do". I wish I could suggest some incontrovertible chain of apodictic reasoning that would overcome this argument, but the best I can do is the following: He has been fairly successful in his efforts to bribe the parasitic with the earnings of the productive. That just goes to show what can be done with a little greed, a complete lack of scruples, and the help of a bunch of insensitive traitors.
That's for broadcast digital TV. Here are the U.S. TV markets, assuming they mean DMA markets (unclear). Everybody bigger than Jackson, TN has to pay (and I'd imagine that Jackson will, too, after the next census data update). This part has nothing to do with Internet streaming. It has to do with the number of unique programs the TV station simultaneously transmits. A station that airs a network affiliate on its first subchannel, 24-hour news on its second subchannel, and 24-hour weather on its third subchannel. If that is a top-172 DMA station, it will cost $30,000 per year to do so because there are three services. It doesn't matter if that spans into a second market using a repeater; it still costs only $30,000. At least that's the way I would interpret it.
Like I said, you have to really get one angry to get bitten. Oh, and most people aren't stupid enough to pick one up in such a way that it can bite you....
No, I'm not misunderstanding you. I'm saying that in the absence of a dedicated bike lane, radical differences between the fastest and slowest vehicles on the road are the most common root cause of accidents that don't involve some form of impairment. Therefore, in the absence of dedicated bike lanes, bicycles cause significant problems with traffic flow, higher accident rates, etc. They may primarily cause problems for cars, but that does not make it a car problem (and when a bicyclist gets hit, it clearly becomes a bicycle problem). It is a compatibility problem. Blaming cars for not being able to share roads well with cars is like blaming a Linux user for not being able to share data easily with an EBCDIC mainframe. They're just massively incompatible technologies.
But seriously, tarantulas may be annoying, but they aren't what most people would call poisonous. This particular one is among the worst, as it can cause several hours of vomiting in humans; it won't kill you, but you might wish you were dead.:-) And they can kill pets. Fortunately, they are also not particularly aggressive towards people. You have to really, really piss off a tarantula to get bitten. We used to pick up tarantulas (not this particular species) and let them crawl around on us as kids. They look scary, but in general if you don't bother them, they won't bother you.
Why? Because the employers are the ones choosing to locate in densely populated areas. The employees aren't choosing to work in those locations. It's either work there or find another job. The onus should be upon companies bringing large number of employees into a city to help pay for the costs of doing business in those cities. You wouldn't expect a shopping mall to be built without adequate parking for customers. Why, then, should an office building be any different? Giving employees a place to park their cars is a part of doing business.
It's the same reason that at least here in California, businesses building high density housing are required to have a certain number of parking spaces per housing unit (I think it is something like 2.5, but I don't remember for sure beyond that it is greater than one). Without such laws, many unscrupulous businesses would cut every corner they could during rental unit construction and the people who rent there would end up finding other places to park, putting added strain on limited street parking and generally creating a public nuisance.
By proactively forcing businesses to be good citizens and take care of parking for their employees, tenants, etc., you avoid the parking problems that places like New York have. In the long run, I think you will find that it is much cheaper to build enough parking to begin with than to try to manage the problems that are caused by not having enough parking where it is needed.
The initial couple of blocks of city streets notwithstanding, that's what a Garmin or Tom Tom is for. The newer ones can get real-time traffic data over FM sidebands and route you around traffic problems before you get to them. My trip to Santa Cruz yesterday would have taken probably a half hour hour longer if it hadn't directed me onto Montague Expressway to route around a backup on 101. Besides, public transit isn't immune to traffic backups, either. Busses still get caught in traffic and can take hours to get where they are going.
$450 per month for insurance is obscene anywhere. The highest average car insurance anywhere in Canada is only $1,400 annually in British Columbia. (Source: CityNews.) You're talking about almost four times that.
IMHO, the fact that you'd have to pay anything to park at work is a sign of shortsightedness on the part of your employer. I work in a company that can house almost all of its employees' cars underneath the building, and that was built before this area started to acquire significant density. IMHO, it is downright criminal to build a corporate office building without providing adequate parking for employees. That's like building a car with a gallon jug for a gas tank. It just makes no sense to build buildings that way in a dense urban area. Underground parking should be a part of the cost of doing business in a city, IMHO---possibly even mandated by law.
Slashdot Mirror
Juice had better not have added sugar. If it does, it isn't juice. Don't be fooled by things that don't say that they are juice (e.g. Kern's "Nectar", which is almost entirely high fructose corn syrup).
As for whether juice is bad for you, a lot of that depends on whether it brings dietary fiber along with it. Opt for pressed/unfiltered apple juice rather than the filtered stuff, for example.
For that matter, Lean Cuisine meals are about $2.00 apiece at Target, and qualify as a reasonably healthy meal. Double that at a grocery, but even still, it's not significantly more expensive than a 1/4 pounder, fries, and a Coke.
That's not quite right. The answers to FB quizzes aren't stored as part of your profile. AFAICT, FB developers don't have access to any storage as part of the FB platform. That's one of the things I dislike about the concept of writing FB apps: it's BYOS (bring your own storage).
Thus, if the results are stored at all, they are stored by the application developer. It may be tied to your profile ID, but it is not part of your profile. Only the text that it spits out at the end is stored as part of your profile, and only if you tell it to add that result to your wall or whatever.
I realize I'm being pedantic here and that it makes little difference in practice, but....
The IQ test is not a Facebook app, and people aren't really sending it to you. That's just an ad styled to look like it is coming from Facebook. Real Facebook apps have access to your phone number from your profile and wouldn't need to ask.
How did I lose a space there? *sigh* "Its ads and style make it look like one." Must be too early in the morning.
No, most of the tests are user-generated things created with one of a couple of "quiz generator" applications. The IQ test, AFAIK, is not a Facebook app, though its ads and stylemake it look like one. It's a scam to get people to give them their cell phone numbers.
Meeesa ruler?
Likewise. Then again, I'm caucasian with only medium-dark hair. Call me cynical, but $10 says that if somebody with dark hair, darker skin and Middle-Eastern features did the same thing, there would be plainclothes security people casually walking by but watching that person really carefully within two minutes.
If you get enough sunlight (or artificial equivalent), you can produce all the vitamin D you need yourself. So this basically boils down to "energy" again. In fact, AFAIK, you basically can't get enough vitamin D from your diet alone unless the foods or drinks you consume have been artificially enriched (e.g. milk). Even if you could get enough from your diet, you'd probably end up massively overdosing on other vitamins. :-)
According to one of the sample headers that ships with HeaderDoc, TrainsCanFly.
@category Trains_2(TrainsCanFly)
@abstract The TrainsCanFly category adds levitation methods to the Train class.
@discussion Methods declared in the TrainsCanFly category of the Train class can only be used with properly equiped Train objects.
*/
@interface Trains_2(TrainsCanFly)
@method levitateToHeight:
@abstract Raises the train specified number of centimeters
@param height The number of centimeters to levitate.
*/
- (void)levitateToHeight:(float)height;
@end
There are actually a fair number of iPhone apps on the store that provide DAV or HTTP servers for downloading files from the phone, so a BitTorrent client isn't exactly infeasible. It wouldn't be particularly useful given the limited amount of storage on the device, of course.... And as others have noted, this is basically just a native UI skin for a web interface to a BitTorrent client on another machine, not an actual client. Still, I'd probably be worried about contributory infringement claims if somebody tried to host such an app on a site I ran, so I can see why Apple would shy away from it.
That said, I'm left wondering why the developers of this application didn't just write it as a web app. If you take advantage of the HTML5 offline application cache to allow it to be cached on disk (flash), there's really not much benefit to writing a native app for something this trivial. That would avoid the whole question of whether Apple would allow it on the app store.
Just my $0.02.
I think I see part of the confusion in this thread. I've read some of the papers and they confused me at first. There are two terms called a "block". There is a logical block (512 bytes) and there is a flash block, which is a group of flash pages that must be erased together. Flash pages can be reordered within flash blocks, at least if you're talking about temporary log blocks. I haven't found any evidence that reordering is done in 512-byte quantities. Technical papers tend to use the word "sector" to refer to a 512-byte block to minimize confusion caused by the massively overloaded term "block".
As an aside, this all points to a need for trained technical writers/editors to be involved in the crafting of technical papers; a trained writer or editor would likely have immediately complained about the reuse of the word "block". Flash is really a painful edge case in that the physical block size is larger than the logical block size, which is just plain wrong conceptually. Someone would have insisted on a more precise name like "page group" instead of "flash block". This confusion should have been nipped in the bud years ago.... :-)
BTW, I *think* (but I'm not certain) that the linked post is confusing block storage with log storage. AFAIK, the only truly fragmented flash blocks are out-of-band log blocks that contain short-term copies of flash pages that were recently written. Eventually, those writes get flushed out to their normal places within standard, page-mapped flash pages. It is basically a highly persistent write cache.
The controller doesn't have to work around that. When those get full, the still-pertinent data in older log blocks has to be flushed, so yes, there's a performance hit---it degrades to the true random write performance of the device, give or take---but not because of fragmentation. The performance hit is because the cache is full. AFAIK, the controller should not generally be erasing these and rewriting them with some blocks replaced. That would defeat the whole purpose of using log blocks in the first place, which is to minimize the number of erasures during writes to improve burst write performance.
I guess you might make the argument that erasing and rewriting it would be faster if most of the cells in the log block are no longer valid (which in real-world use would probably mostly involve metadata blocks that get rewritten frequently), and I'd imagine that some controllers do perform that sort of tradeoff analysis, but I would expect that in anything approaching a truly random write test, the controller would be flushing far more often than consolidating unless your OS disk cache is really dumb/nonexistent. :-)
Unless I'm misunderstanding things, of course, which is entirely possible....
Yes, I know that extra NAND storage is traditionally used for the mapping. I was initially thinking this would increase write amplification by an order of magnitude or more until the device was full. I have since thought through this a few dozen more times and convinced myself that this could be avoided.
Come again? I'm not aware of SSDs doing any mapping at that level of granularity; to do so would mean that a 256 GB hard drive would require something like 2 GB of storage (assuming an even 32-bit LBA) just to hold the sector mapping table, and that would have to be NOR flash or single-level NAND flash, making it really freakishly expensive.
All SSDs that I'm aware of work like this: you have a logical block address (sector number). These are grouped (contiguously) into pages (those 128KB "slices" you're referred to). If you have a 128 MB page size, then sectors 0-255 would be logical page 0, sectors 1-511 are logical page 1, and so on.
These logical pages are then arbitrarily mapped to physical pages. This results in a mapping table that is a much more plausible 8 megabytes in size for a 256 GB flash drive.
Thus, any fragmentation within a flash page is caused entirely by the filesystem being fragmented, not by anything happening in the flash controller. The performance degradation from full flash has nothing to do with fragmentation. It is caused by the flash controller having to erase an already-used page before it writes to it. The only way to avoid that is to erase unmapped spare pages ahead of time.
No, when people get screwed by it calling a legitimate installation bogus, they'll say they got a WAT the f*ck.
Wow. I ran that on President Bush, and the first paragraph was actually pretty close to comprehensible and on-target....
That's for broadcast digital TV. Here are the U.S. TV markets, assuming they mean DMA markets (unclear). Everybody bigger than Jackson, TN has to pay (and I'd imagine that Jackson will, too, after the next census data update). This part has nothing to do with Internet streaming. It has to do with the number of unique programs the TV station simultaneously transmits. A station that airs a network affiliate on its first subchannel, 24-hour news on its second subchannel, and 24-hour weather on its third subchannel. If that is a top-172 DMA station, it will cost $30,000 per year to do so because there are three services. It doesn't matter if that spans into a second market using a repeater; it still costs only $30,000. At least that's the way I would interpret it.
Like I said, you have to really get one angry to get bitten. Oh, and most people aren't stupid enough to pick one up in such a way that it can bite you....
No, I'm not misunderstanding you. I'm saying that in the absence of a dedicated bike lane, radical differences between the fastest and slowest vehicles on the road are the most common root cause of accidents that don't involve some form of impairment. Therefore, in the absence of dedicated bike lanes, bicycles cause significant problems with traffic flow, higher accident rates, etc. They may primarily cause problems for cars, but that does not make it a car problem (and when a bicyclist gets hit, it clearly becomes a bicycle problem). It is a compatibility problem. Blaming cars for not being able to share roads well with cars is like blaming a Linux user for not being able to share data easily with an EBCDIC mainframe. They're just massively incompatible technologies.
Or flame throwers, napalm, etc.
But seriously, tarantulas may be annoying, but they aren't what most people would call poisonous. This particular one is among the worst, as it can cause several hours of vomiting in humans; it won't kill you, but you might wish you were dead. :-) And they can kill pets. Fortunately, they are also not particularly aggressive towards people. You have to really, really piss off a tarantula to get bitten. We used to pick up tarantulas (not this particular species) and let them crawl around on us as kids. They look scary, but in general if you don't bother them, they won't bother you.
Bikes flow through traffic jams, but they would clog major highways hopelessly if they were allowed there. With every choice, there are tradeoffs.
Why? Because the employers are the ones choosing to locate in densely populated areas. The employees aren't choosing to work in those locations. It's either work there or find another job. The onus should be upon companies bringing large number of employees into a city to help pay for the costs of doing business in those cities. You wouldn't expect a shopping mall to be built without adequate parking for customers. Why, then, should an office building be any different? Giving employees a place to park their cars is a part of doing business.
It's the same reason that at least here in California, businesses building high density housing are required to have a certain number of parking spaces per housing unit (I think it is something like 2.5, but I don't remember for sure beyond that it is greater than one). Without such laws, many unscrupulous businesses would cut every corner they could during rental unit construction and the people who rent there would end up finding other places to park, putting added strain on limited street parking and generally creating a public nuisance.
By proactively forcing businesses to be good citizens and take care of parking for their employees, tenants, etc., you avoid the parking problems that places like New York have. In the long run, I think you will find that it is much cheaper to build enough parking to begin with than to try to manage the problems that are caused by not having enough parking where it is needed.
The initial couple of blocks of city streets notwithstanding, that's what a Garmin or Tom Tom is for. The newer ones can get real-time traffic data over FM sidebands and route you around traffic problems before you get to them. My trip to Santa Cruz yesterday would have taken probably a half hour hour longer if it hadn't directed me onto Montague Expressway to route around a backup on 101. Besides, public transit isn't immune to traffic backups, either. Busses still get caught in traffic and can take hours to get where they are going.
$450 per month for insurance is obscene anywhere. The highest average car insurance anywhere in Canada is only $1,400 annually in British Columbia. (Source: CityNews.) You're talking about almost four times that.
IMHO, the fact that you'd have to pay anything to park at work is a sign of shortsightedness on the part of your employer. I work in a company that can house almost all of its employees' cars underneath the building, and that was built before this area started to acquire significant density. IMHO, it is downright criminal to build a corporate office building without providing adequate parking for employees. That's like building a car with a gallon jug for a gas tank. It just makes no sense to build buildings that way in a dense urban area. Underground parking should be a part of the cost of doing business in a city, IMHO---possibly even mandated by law.