Still, even in the best case, there's still basically an Amazon-eBay duopoly. Most people never even see anything else. So in terms of sales volume, if you aren't on one of those or both, you're going to be in a world of hurt. And the bigger concern is that these sorts of agreements tend to be signs of things to come; there's nothing preventing Apple from entering into a similar agreement with eBay.
So, what you are saying is, that you have SELF-SELECTED Amazon, and then have the temerity to call it a "Monopoly"?
What I am saying is that a very large percentage of consumers have self-selected Amazon as being the only trustworthy means of buying things from third-party sellers with adequate consumer protection, and that they in effect have no viable competition. More than 75% of online shoppers say that they shop at Amazon most of the time.
People are concerned indeed... but not in the direction you think. They are concerned about getting authorized refurb units, not from some fly-by-night yahoo selling unknown crap out of his basement.
Uh, no. You don't understand the market. Apple sells refurb units of a given model for maybe a year, typically. These are mostly units that had serious problems right out of the box, during the period of high-volume sales shortly after the product comes out, that got replaced by Apple under warranty. After the first year, you typically do not see Apple selling that model in refurb form, because in the rare event that they have to replace one after the first year, the dud machine gets parted out and used to repair other people's machines.
By contrast, nobody scraps a machine during that time period, because the products are under warranty, and even in the case of abuse, repair is still almost always going to be much cheaper than outright replacement. Therefore, these machines are almost invariably going to be from prior model years.
Thus, the authorized Apple refurb market and the third-party refurb market are pretty much entirely non-overlapping. The former is a new, current-model product with a full Apple warranty, and the latter is a used product from a previous generation with a third-party warranty.
People who buy used products don't care about Apple refurbs, because they are trying to save money, and you can't save enough money that way to really matter much. And people who don't buy used products won't buy the third-party refurbs, because they want the current model. Speaking as someone who has done both at different points in my life, your argument is nonsense.
In the internet sales business, Amazon has effectively become a monopoly.
Oh, right, because there are NO other companies to buy from "in the internet sales business".
None that anyone buys from. The mere existence of another store does not meet the minimum requirements for competition. There must also be a reasonable possibility of actually selling goods there.
As a buyer, calling Amazon a monopoly is pretty accurate. Apart from a very small number of product-line-specific resellers that I buy from, the only other selling platform out there is eBay, which is basically only a last resort if I can't find what I want on Amazon. More to the point, not once have I bought anything on eBay merely because it was cheaper there. I don't even *start* looking at eBay unless Amazon has failed to provide the product. So at least for me, sellers that sell solely on eBay are not competing with Amazon in any meaningful way.
But even that is not what makes this disconcerting. With this policy, Amazon is restricting not just new product sales, but also used product sales. After all, that's what a refurbished product is. Never before in the history of mankind have two companies with as much market power as Apple and Amazon colluded to destroy the used market for their products. Anyone who owns an Apple product should be very, very concerned, both with the realization that Apple wants them to be unable to sell their products when they want to upgrade and with the horrible net impact on the environment resulting from such policies.
Shame on you, Apple. And shame on you, Amazon, for going along with it. I have never been more disappointed in Apple in all my life.
If it should have never been asked or was inappropriate, he or his lawyers should have said so at the time.
Yes, they should have, because it was immaterial. Further, legally speaking, he did not commit perjury, because it was immaterial.
Why was it immaterial? Because all evidence strongly indicates that Lewinsky initiated the relationship, not Clinton. Clinton's alleged behavior, as described by Jones, was so dramatically different than what occurred in the Lewinsky relationship that no evidence related to the latter relationship was even remotely relevant to the Jones case, per Federal Rule of Evidence 401 (regardless of whether you look at the pre-2011 or post-2011 version of that rule).
Ergo, lie or not, he did not commit perjury, and the Senate was correct to not convict him.
In Seattle, the numbers for these "Migrant Homeless" persons was in the 5-10% range. The other 90% were split among persons who lived in the county prior to being homeless (most) and other places in the state (few).
Let's assume that the 90% number also holds for everywhere else on the west coast and put that into perspective. Almost 16% of the population lives on the coast to begin with, so if we assume that:
Homeless are distributed equally
The east coast already has strong homeless services, and thus their homeless probably don't leave
People tend to go to each coast in equal numbers
then we would expect about 16% on the west coast and about 15% or a little over on the east coast, leaving only 69% of the homeless distributed across the rest of the country, and if half of them go to each coast, then maybe 35% are likely to move to the west coast rather than the east coast.
Now if 10% of the homeless are from somewhere else, divide 16% by.9 and you would then expect 17.8 percent to be in California, rather than being evenly distributed. Subtract the original 16%, and this suggests that about 1.8% of the U.S. homeless population has moved to California, out of a likely 35%. So likely one out of every twenty non-coastal homeless people have moved to one coast or the other.
I know there are a lot of assumptions in there, some of which could be wrong, and in particular, the assumption that homeless people don't leave the coasts isn't a given, though the equal distribution part is so conservative(*) that it probably more than cancels out the exceptions to the other two assumptions. Either way, the fact that only 10% of the homeless on the coasts are from other areas doesn't even remotely disprove the theory that homeless people move to the coasts because of weather or available homeless services. In fact, to me, it suggests that doing so is probably a lot more common than I would otherwise have assumed.
(*) Urban homelessness, AFAIK, is a lot more common than rural homelessness, so you would expect more homeless near the relatively urban coasts than in the rest of the country even before you factor in any people who decide to move to California after becoming homeless.
You're just talking about the health-related deaths. I was talking about traffic deaths, where there is a statistically significant increase at BOTH time changes — specifically, on the Monday after the spring shift to DST and, curiously, on the Sunday of the fall shift away from DST.
He lied in response to a question that should never have even been asked. The entire line of questioning was inappropriate, as was the investigation leading up to it, because a consensual relationship (and at no point was there even the slightest hint that the relationship was anything less than consensual) is simply not a legitimate reason to investigate a sitting POTUS, married or not. Don't get me wrong, what he did was wrong morally and ethically, but it wasn't criminal, and initiating such a highly inappropriate and public investigation in a deliberate attempt to put someone into a position where embarrassment would cause that person to lie under oath so that you can convict the person of a crime is pretty much the very definition of entrapment.
So Clinton was morally flawed (and really, who isn't), but the Republican leaders were morally bankrupt.
It is to maximize sunlight hours when people aren't at work or school. Yes morning commutes will become more hazardous but you'll have more nature light for whatever outdoor activities after work.
And evening commutes will become much, much less hazardous. People can easily use artificial lighting to wake themselves up in the morning for their morning drives. But at the end of a long day, when they get in the car for their evening drives, they're tired, so darkness has a much bigger impact. Thus, I would expect a significantly larger reduction in traffic deaths from moving to year-round DST when compared with moving to year-round ST.
Of course either approach is better than the two days of carnage that we get under the current scheme.
As for impeaching Trump, since Republicans hold the Senate, that's not going to happen.
That's not how impeachment works. The House impeaches by itself. The Senate convicts. Without the conviction, the impeachment has no effect. That's why Bill Clinton was impeached, but not removed from office.
So if the Democrats were as petty as the Republicans, they could impeach Trump. They do not have the power to remove him from office unless they can prove wrongdoing sufficient to convince some of the Republicans in the Senate to convict him, but they could easily make a show of it and repeat the appalling behavior that we saw back in the Clinton era that sent our government completely off the rails.
Send it where? By mail? If we were designing a new phone system, we'd do it as a packet-switched digital network and have fields in the ring packet for that. The POTS is a circuit switched network. There are no fields, much less an easy way to add new fields.
As I said in the original post, extend the CID/CND protocol. The caller ID message starts with a byte that tells that it is a caller ID blob. I'm assuming that not all 256 possible codes are currently used, so you could send a second blob right after or before it with a different code that contains a caller ID authentication token, and specify a different message type code.
If we were designing the phone system from scratch, if we didn't already have a phone system, it would be easy enough to include signed token using PKI. The simplest approach, defining exactly how the encryption works, would be cracked in a few years, so really we'd want a framework for negotiating the encryption, similar to ipsec. That would certainly be doable.
Meh. The simplest approach is for the owning phone company to generate a random 128-bit number for each call, store it in a database along with a time stamp, and send the time stamp and random number. This approach can never be "cracked", so long as the phone company deletes the entry when the call is completed.
The actual contents of the token, including how to verify it, should be left as an implementation detail for the company that owns the phone number. As far as everyone else along the line is concerned, it is an opaque authentication token, and when they ask the owning company whether it is valid or not, they get back a boolean.
This could be easily solved, of course. Just extend the CID protocol to include a timestamped, signed blob from the original provider that can be passed on by any forwarders and verified with the original provider for the duration of the call.
Or, for that matter, assuming your upstream provider knows what's going on, they could ostensibly just whitelist incoming calls' numbers for outgoing forwarding as soon as they come in, and remove that whitelisting as soon as the incoming call goes away.
But in an urban environment, they are kind of useless, especially the high speed kind, because a significant number of folks only want to go part way between A and B, and the tunnel is worthless for them.
First, your assumption that a tunnel cannot have multiple entrances and exits is wrong. I've been in several underground tunnels in which I had to make an exit decision while underground if I wanted to end up in the right place. (This tents to wreak havoc with GPS, but that's a solvable and somewhat orthogonal problem.)
Second, even if you were correct, the tunnel still wouldn't be worthless for people who only need to go part of that distance. All of the cars that are going from anywhere near A to anywhere near B are likely to take the tunnel because it is much faster. Every car that does so represents a reduction in traffic on the main surface highways and streets for the entire stretch of highway that they bypassed, which means those roads also get faster because of the reduced traffic volume.
The key is to choose endpoints that cover stretches in which a significant percentage of traffic is going all the way. For example, a bypass under CA-101 that skips all of Palo Alto and Mountain View would be a huge win even for the Google employees getting on at Mountain View, because of all the traffic that they wouldn't have to merge into at 5 MPH.
An ideal implementation would likely involve running tunnels in parallel with existing surface highways, with the surface highways retaining all of their existing exits, and the tunnels allowing entrance and exit every several exits, thus reducing the frequency of merges that currently cause slowdowns for vehicles traveling long distances, and making the merges faster for people who just going a short distance. (Incidentally, this is also how you improve the speed of mass transit.)
Which means they have more fat, which means higher estrogen, which translates to less sadistic behavior. So perhaps drinking black coffee causes sadism?
Being a landlord is not generally considered to be rent seeking. Rent seeking means seeking to make profits from something that you already have created without contributing anything back to society. A landlord contributes the use of property to society in exchange for that money. There is an actual cost to the landlord, both in terms of being unable to use that property for other things and in terms of having to maintain that property.
It could be argued that the iOS App Store qualifies as rent seeking — not because there isn't a cost to Apple associated with providing the store, but because Apple uses policies that prevent users from side-loading apps or using other app stores to enforce their ability to make profits on all apps used on the platform, and some people would argue that there is no benefit to society from those road blocks. (This, of course, boils down to the whole closed versus open ecosystem debate.)
That was my thought as well, but for a different reason. In my mind, there are two situations when I use a map. One is while driving. The other is while figuring out an area that I'm trying to learn about for some reason.
While driving, I want the fewest distractions possible. Beyond street shape and street names, every additional piece of visual information competing for my attention is more likely to be a distraction than a helpful hint. Maybe showing the names of businesses that are in the same category as those I've recently searched for might be useful. Maybe. And that's right at the limit of things that would have any utility whatsoever. Something as meaningless as whether a median has grass or not is complete noise that just makes the text harder to read.
When I'm learning about an area, I want as much visual information as possible, which means things like satellite imagery, street views, etc.
In no case have I ever thought, "I'd like something halfway between these two extremes, where I can see the greenness of the grass, but without being able to see that it is grass."
The problem with Apple Maps has always been, in my opinion, that it was designed with a "vector images are neat; let's see how visually pleasing a map we can draw with them" approach, rather than a "let's figure out how users use maps, and design an optimal solution for real-world use cases" approach.
For example, six years after Apple Maps users first started filing bugs complaining that Apple Maps gives you no way to tell the difference between "there is no traffic data on this road" and "traffic is moving freely on this road", this fundamental usability flaw is still not fixed. (And don't get me started on Tesla, who just started pulling the same crap with their maps.)
IMO, Apple needs to rethink Apple Maps from the ground up, focusing on how people use maps, and completely forgetting about how the maps look. Make it work, then make it pretty. As long as you're trying to do it the other way around, it will always be pretty, and will never really work.
Well, yes and no. I've been dealing with it on my iPhone 6s ever since iOS 9 was released. It usually happens at 10% or lower. Every now and then, it will fail at fifteen percent or so. I make it a point never to let it get below 50%.
Yes, in colder temperatures, batteries produce less power, so it could happen at a higher level of charge, but I'd be shocked if it happened on a battery that was more than half full, even at crazy low temperatures.
Ignoring that skepticism, this is the difference between the phone dying at 0% and maybe surprising you by shutting down at a few percent instead of at zero. If you're doing something that is critical to finish, you shouldn't be doing it on a nearly empty battery.
Because prior to the next model release, every device from the previous generation is under warranty, give or take a couple of weeks, so battery replacements would be free.
During the warranty period, if a device exhibits sudden shutdowns, it is better for the customer if that device prematurely shuts down repeatedly, because that encourages the customer to get his or her battery replaced rather than continue to suffer from a bad battery that could have been fixed for free.
Slashdot Mirror
Not in the United States.
Still, even in the best case, there's still basically an Amazon-eBay duopoly. Most people never even see anything else. So in terms of sales volume, if you aren't on one of those or both, you're going to be in a world of hurt. And the bigger concern is that these sorts of agreements tend to be signs of things to come; there's nothing preventing Apple from entering into a similar agreement with eBay.
What I am saying is that a very large percentage of consumers have self-selected Amazon as being the only trustworthy means of buying things from third-party sellers with adequate consumer protection, and that they in effect have no viable competition. More than 75% of online shoppers say that they shop at Amazon most of the time.
Uh, no. You don't understand the market. Apple sells refurb units of a given model for maybe a year, typically. These are mostly units that had serious problems right out of the box, during the period of high-volume sales shortly after the product comes out, that got replaced by Apple under warranty. After the first year, you typically do not see Apple selling that model in refurb form, because in the rare event that they have to replace one after the first year, the dud machine gets parted out and used to repair other people's machines.
By contrast, nobody scraps a machine during that time period, because the products are under warranty, and even in the case of abuse, repair is still almost always going to be much cheaper than outright replacement. Therefore, these machines are almost invariably going to be from prior model years.
Thus, the authorized Apple refurb market and the third-party refurb market are pretty much entirely non-overlapping. The former is a new, current-model product with a full Apple warranty, and the latter is a used product from a previous generation with a third-party warranty.
People who buy used products don't care about Apple refurbs, because they are trying to save money, and you can't save enough money that way to really matter much. And people who don't buy used products won't buy the third-party refurbs, because they want the current model. Speaking as someone who has done both at different points in my life, your argument is nonsense.
In the internet sales business, Amazon has effectively become a monopoly.
Oh, right, because there are NO other companies to buy from "in the internet sales business".
None that anyone buys from. The mere existence of another store does not meet the minimum requirements for competition. There must also be a reasonable possibility of actually selling goods there.
As a buyer, calling Amazon a monopoly is pretty accurate. Apart from a very small number of product-line-specific resellers that I buy from, the only other selling platform out there is eBay, which is basically only a last resort if I can't find what I want on Amazon. More to the point, not once have I bought anything on eBay merely because it was cheaper there. I don't even *start* looking at eBay unless Amazon has failed to provide the product. So at least for me, sellers that sell solely on eBay are not competing with Amazon in any meaningful way.
But even that is not what makes this disconcerting. With this policy, Amazon is restricting not just new product sales, but also used product sales. After all, that's what a refurbished product is. Never before in the history of mankind have two companies with as much market power as Apple and Amazon colluded to destroy the used market for their products. Anyone who owns an Apple product should be very, very concerned, both with the realization that Apple wants them to be unable to sell their products when they want to upgrade and with the horrible net impact on the environment resulting from such policies.
Shame on you, Apple. And shame on you, Amazon, for going along with it. I have never been more disappointed in Apple in all my life.
Yes, they should have, because it was immaterial. Further, legally speaking, he did not commit perjury, because it was immaterial.
Why was it immaterial? Because all evidence strongly indicates that Lewinsky initiated the relationship, not Clinton. Clinton's alleged behavior, as described by Jones, was so dramatically different than what occurred in the Lewinsky relationship that no evidence related to the latter relationship was even remotely relevant to the Jones case, per Federal Rule of Evidence 401 (regardless of whether you look at the pre-2011 or post-2011 version of that rule).
Ergo, lie or not, he did not commit perjury, and the Senate was correct to not convict him.
Let's assume that the 90% number also holds for everywhere else on the west coast and put that into perspective. Almost 16% of the population lives on the coast to begin with, so if we assume that:
then we would expect about 16% on the west coast and about 15% or a little over on the east coast, leaving only 69% of the homeless distributed across the rest of the country, and if half of them go to each coast, then maybe 35% are likely to move to the west coast rather than the east coast.
Now if 10% of the homeless are from somewhere else, divide 16% by .9 and you would then expect 17.8 percent to be in California, rather than being evenly distributed. Subtract the original 16%, and this suggests that about 1.8% of the U.S. homeless population has moved to California, out of a likely 35%. So likely one out of every twenty non-coastal homeless people have moved to one coast or the other.
I know there are a lot of assumptions in there, some of which could be wrong, and in particular, the assumption that homeless people don't leave the coasts isn't a given, though the equal distribution part is so conservative(*) that it probably more than cancels out the exceptions to the other two assumptions. Either way, the fact that only 10% of the homeless on the coasts are from other areas doesn't even remotely disprove the theory that homeless people move to the coasts because of weather or available homeless services. In fact, to me, it suggests that doing so is probably a lot more common than I would otherwise have assumed.
(*) Urban homelessness, AFAIK, is a lot more common than rural homelessness, so you would expect more homeless near the relatively urban coasts than in the rest of the country even before you factor in any people who decide to move to California after becoming homeless.
You're just talking about the health-related deaths. I was talking about traffic deaths, where there is a statistically significant increase at BOTH time changes — specifically, on the Monday after the spring shift to DST and, curiously, on the Sunday of the fall shift away from DST.
He lied in response to a question that should never have even been asked. The entire line of questioning was inappropriate, as was the investigation leading up to it, because a consensual relationship (and at no point was there even the slightest hint that the relationship was anything less than consensual) is simply not a legitimate reason to investigate a sitting POTUS, married or not. Don't get me wrong, what he did was wrong morally and ethically, but it wasn't criminal, and initiating such a highly inappropriate and public investigation in a deliberate attempt to put someone into a position where embarrassment would cause that person to lie under oath so that you can convict the person of a crime is pretty much the very definition of entrapment.
So Clinton was morally flawed (and really, who isn't), but the Republican leaders were morally bankrupt.
And evening commutes will become much, much less hazardous. People can easily use artificial lighting to wake themselves up in the morning for their morning drives. But at the end of a long day, when they get in the car for their evening drives, they're tired, so darkness has a much bigger impact. Thus, I would expect a significantly larger reduction in traffic deaths from moving to year-round DST when compared with moving to year-round ST.
Of course either approach is better than the two days of carnage that we get under the current scheme.
That's not how impeachment works. The House impeaches by itself. The Senate convicts. Without the conviction, the impeachment has no effect. That's why Bill Clinton was impeached, but not removed from office.
So if the Democrats were as petty as the Republicans, they could impeach Trump. They do not have the power to remove him from office unless they can prove wrongdoing sufficient to convince some of the Republicans in the Senate to convict him, but they could easily make a show of it and repeat the appalling behavior that we saw back in the Clinton era that sent our government completely off the rails.
As I said in the original post, extend the CID/CND protocol. The caller ID message starts with a byte that tells that it is a caller ID blob. I'm assuming that not all 256 possible codes are currently used, so you could send a second blob right after or before it with a different code that contains a caller ID authentication token, and specify a different message type code.
No, no, today is *Independent* voting day. Republican day was yesterday. Sorry you missed it. Better luck in two years. :-D
Meh. The simplest approach is for the owning phone company to generate a random 128-bit number for each call, store it in a database along with a time stamp, and send the time stamp and random number. This approach can never be "cracked", so long as the phone company deletes the entry when the call is completed.
The actual contents of the token, including how to verify it, should be left as an implementation detail for the company that owns the phone number. As far as everyone else along the line is concerned, it is an opaque authentication token, and when they ask the owning company whether it is valid or not, they get back a boolean.
This could be easily solved, of course. Just extend the CID protocol to include a timestamped, signed blob from the original provider that can be passed on by any forwarders and verified with the original provider for the duration of the call.
Or, for that matter, assuming your upstream provider knows what's going on, they could ostensibly just whitelist incoming calls' numbers for outgoing forwarding as soon as they come in, and remove that whitelisting as soon as the incoming call goes away.
First, your assumption that a tunnel cannot have multiple entrances and exits is wrong. I've been in several underground tunnels in which I had to make an exit decision while underground if I wanted to end up in the right place. (This tents to wreak havoc with GPS, but that's a solvable and somewhat orthogonal problem.)
Second, even if you were correct, the tunnel still wouldn't be worthless for people who only need to go part of that distance. All of the cars that are going from anywhere near A to anywhere near B are likely to take the tunnel because it is much faster. Every car that does so represents a reduction in traffic on the main surface highways and streets for the entire stretch of highway that they bypassed, which means those roads also get faster because of the reduced traffic volume.
The key is to choose endpoints that cover stretches in which a significant percentage of traffic is going all the way. For example, a bypass under CA-101 that skips all of Palo Alto and Mountain View would be a huge win even for the Google employees getting on at Mountain View, because of all the traffic that they wouldn't have to merge into at 5 MPH.
An ideal implementation would likely involve running tunnels in parallel with existing surface highways, with the surface highways retaining all of their existing exits, and the tunnels allowing entrance and exit every several exits, thus reducing the frequency of merges that currently cause slowdowns for vehicles traveling long distances, and making the merges faster for people who just going a short distance. (Incidentally, this is also how you improve the speed of mass transit.)
It wants to be closer to your jugular. Your cat is biding its time.
Which means they have more fat, which means higher estrogen, which translates to less sadistic behavior. So perhaps drinking black coffee causes sadism?
More like saying, "I'm a C programmer because I write iOS apps (in Objective-C)."
Being a landlord is not generally considered to be rent seeking. Rent seeking means seeking to make profits from something that you already have created without contributing anything back to society. A landlord contributes the use of property to society in exchange for that money. There is an actual cost to the landlord, both in terms of being unable to use that property for other things and in terms of having to maintain that property.
It could be argued that the iOS App Store qualifies as rent seeking — not because there isn't a cost to Apple associated with providing the store, but because Apple uses policies that prevent users from side-loading apps or using other app stores to enforce their ability to make profits on all apps used on the platform, and some people would argue that there is no benefit to society from those road blocks. (This, of course, boils down to the whole closed versus open ecosystem debate.)
Use both sides of the train.
That was my thought as well, but for a different reason. In my mind, there are two situations when I use a map. One is while driving. The other is while figuring out an area that I'm trying to learn about for some reason.
While driving, I want the fewest distractions possible. Beyond street shape and street names, every additional piece of visual information competing for my attention is more likely to be a distraction than a helpful hint. Maybe showing the names of businesses that are in the same category as those I've recently searched for might be useful. Maybe. And that's right at the limit of things that would have any utility whatsoever. Something as meaningless as whether a median has grass or not is complete noise that just makes the text harder to read.
When I'm learning about an area, I want as much visual information as possible, which means things like satellite imagery, street views, etc.
In no case have I ever thought, "I'd like something halfway between these two extremes, where I can see the greenness of the grass, but without being able to see that it is grass."
The problem with Apple Maps has always been, in my opinion, that it was designed with a "vector images are neat; let's see how visually pleasing a map we can draw with them" approach, rather than a "let's figure out how users use maps, and design an optimal solution for real-world use cases" approach.
For example, six years after Apple Maps users first started filing bugs complaining that Apple Maps gives you no way to tell the difference between "there is no traffic data on this road" and "traffic is moving freely on this road", this fundamental usability flaw is still not fixed. (And don't get me started on Tesla, who just started pulling the same crap with their maps.)
IMO, Apple needs to rethink Apple Maps from the ground up, focusing on how people use maps, and completely forgetting about how the maps look. Make it work, then make it pretty. As long as you're trying to do it the other way around, it will always be pretty, and will never really work.
Well, yes and no. I've been dealing with it on my iPhone 6s ever since iOS 9 was released. It usually happens at 10% or lower. Every now and then, it will fail at fifteen percent or so. I make it a point never to let it get below 50%.
Yes, in colder temperatures, batteries produce less power, so it could happen at a higher level of charge, but I'd be shocked if it happened on a battery that was more than half full, even at crazy low temperatures.
Important work? On an iPhone?
Ignoring that skepticism, this is the difference between the phone dying at 0% and maybe surprising you by shutting down at a few percent instead of at zero. If you're doing something that is critical to finish, you shouldn't be doing it on a nearly empty battery.
Because prior to the next model release, every device from the previous generation is under warranty, give or take a couple of weeks, so battery replacements would be free.
During the warranty period, if a device exhibits sudden shutdowns, it is better for the customer if that device prematurely shuts down repeatedly, because that encourages the customer to get his or her battery replaced rather than continue to suffer from a bad battery that could have been fixed for free.