I didn't know it had to be activated on a computer. I do know there are people who own and use iPhones and do not own personal computers. If there's some initial activation, couldn't they just do that on a computer that they borrow?
In one of the movies -- the third one, I think -- Magneto gives a moving speech, the narrative purpose of which I believe was to shake up the audience's assumptions about which side had just the greater claim to justice.
Then, Magneto ordered one of his followers to kill an innocent person.
I've seen this sort of thing elsewhere, and I wonder if there's a term for it on TVTropes: when there's a strong suggestion that the villain isn't really a villain after all, and then this suggestion is immediately subverted by the villain doing something horrible without any justification.
Come to think of it, I remember comparing notes with other people about the first movie, and nearly everyone agreed that Magneto's plan to turn anti-mutant political leaders into mutants actually seemed like a good idea, but it was also subverted by his kidnapping and torture of Rogue.
On reflection, I understated the significance of ATMs using two-factor authentication. That it's two distinct tests probably matters much more than how strong a password a PIN is.
I'm beginning to think there's excessive paranoia about a very narrow conception of security.
I am asked to invent at least one password a day. Most often, it's for something for which there is no need for any security. In fact, given that users, who are asked to make up passwords frequently, at a moment's notice, understandably reuse passwords, requiring a password where it isn't really necessary actually undermines security.
Every bank I've used, and as far as I know, every bank, requires a four-digit PIN for ATM access. That means that the single most important password most people use daily is a very weak one -- offset by the second form of authentication, the ATM card. ATM cards are frequently left in ATM machines by mistake. How much time do people spend worrying about the security of their bank PINs?
There's an issue going unaddressed here: the idea of a "superhero" is based upon the idea of a "hero," and that is an explicitly aristocratic and anti-egalitarian idea.
The classic hero is a legendary figure who is also supposed to be a group's ancestor, and aristocrats are traditionally direct descendants of heroes. Take the heroes in the Iliad and Odyssey, for example: they are explicitly aristocrats, the ancestors of aristocratic lineages, and they are specially connected to gods and have abilities that set them beyond ordinary people.
The egalitarian ideal posits that people don't vary that much in ability. Not, of course, that everyone is actually equal in ability, but that people don't vary so much that one person should be absolutely elevated over another. When one person has significantly more power than another, it's a morally perilous situation; the person with more power is expected to express humility and restraint.
For every Einstein, there are hundreds of thousands of graduate students who understand Einstein's theories, millions of college students who understand them in outline, and countless high school students who have a cursory understanding of energy-matter equivalence and the twin paradox. Shakespeare had many contemporaries, scarcely known now except by students of drama and literary history, who wrote plays with similar themes and styles. While there are differences in abilities, we tend to oversimplify history and exaggerate those differences.
The egalitarian ideal is based upon the defeat of the aristocratic ideal -- superhero stories, like a lot of fantasy fiction, include aristocratic ideals trying to slip back in.
Stories about "superheroes" have premises that make for ongoing internal and external conflicts: how can you reconcile egalitarian ideals with the existence of individuals with extraordinary power? Thus, the general insistence on secret identities, Superman's discipline of self-restraint, Spiderman's ethical agonizing, the X-Men's struggles with democracy and its subversion. What doesn't seem to be acknowledged is that the egalitarian ideal has a fundamental premise that (super)heroes don't really exist. Consequently, taking the superhero stories literally, the struggles of "good" superheroes to maintain ideals that are premised upon their non-existence fundamentally don't make sense. They only make sense as storytelling because the "good" superheroes keep struggling to support our reality, not their own.
So, yes, if someone is asked to imagine being a superhero, they're imagining themselves as figures that are presumed not to exist, in a real world context -- they're imagining themselves as beings that, by existing, break established social rules. Thus, the choice is between villainy, and a hidden aristocracy -- the latter not actually making much sense. But, imagining a world in which superheroes actually exist, in a self-consistent way, means imagining a world in which aristocratic ideals are valid -- which immediately strikes most people as loathsome.
I'm a little chagrined, as I should have been conscious of the distinctions.
Anyway, I checked the man page and tested the options. "sudo -s" will inherit the environment from the user who executes it -- i.e., use your path, stay in the directory where you executed "sudo -s", etc. So would "sudo bash".
"sudo -i" will simulate logging in as another user -- by default, root -- and so instead of inheriting the environment, will execute the login configuration files for that user.
So, "sudo -i" is a better equivalent to "su -", and "sudo -s" would be the one to use if you want to use the same environment as the regular user.
It's well known that for message boards, blogs, email lists, etc., most readers are "lurkers," who read without posting responses. I expect the metrics vary, but I remember someone's general rule was one hundred lurkers for every active participant in a discussion.
So, I read these stats as meaning that there's an astonishingly high level of response on Twitter -- although that's offset from the comments that there are people gaming Twitter to generate "hype."
Granted, there's the introductory video from Twitter that suggests you use Twitter in just that way.
However, almost everyone I know who uses Twitter uses it primarily to post links to articles or interesting Websites, along with a short comment on the content -- hence the wide use of URL shorteners.
You can enable the root account in Ubuntu if you really want, but it's recommended against. For that matter, logging straight in as root on any *nix system is generally considered a bad idea, security-wise -- standard practice on other systems is to log in as a normal user and use "su -" to switch to a root shell. You definitely shouldn't log straight in as root in a GUI.
Ubuntu (like OS X) uses sudo, and in practice, in the GUI, it means that a user that is part of the admin group is prompted for a password when executing an admin function, like installing a package system-wide, which is a pretty reasonable lower limit for security. At the command line, if you want a root shell, you could just do "sudo -s", which is effectively the same as "su -" on other *nix systems.
Oh, and if you want to run Java plugins, just install "icedtea6-plugin" from Synaptic, or from the command line with "sudo apt-get install icedtea6-plugin".
It's been in the Ubuntu repositories since at least 8.04. It used to be a bit buggy, but it got markedly more reliable a few distributions ago: 9.04 or 9.10, I think.
Every time I ride the bus, I am struck by how many people are using iPhones -- particularly the poor people. It's counter-intuitive, until you realize that if you can only afford one Internet-capable device, instead of several, you're going to choose the one that does the most, which is most likely an iPhone.
I think a more accurate picture is that Microsoft is challenged on the enterprise end by Linux, on the consumer end by Apple, and in the Internet by Google. Microsoft is powerful, but boxed in. I think there are signs that Microsoft is exploring alliances with the open source community in order to break out of that box.
While it's easier to network Windows desktops with Linux servers than it is to network Linux desktops with Windows servers, it's still easier to network Linux desktops with Linux servers. IT staffs tend to prefer homogeneous environments. If Linux servers continue to be preferred, there will be some pressure to shift from Windows desktops to Linux desktops -- which is where the end-of-life of Windows XP is relevant.
Many people are unsatisfied with OpenOffice, but I don't hear a lot of praise for the most recent version of Microsoft Office, either, and that's the killer app for Windows in the workplace. At some point, if OpenOffice or some other office suite breaks through into general use, there will be a much stronger challenge to Window's dominance on the desktop.
I'm using a computer that's about six years old, and I dual-boot Ubuntu 10.10 and Windows 7. Both work well. On Windows, I have trouble with newer games, but that's to be expected; I find Windows 7 runs faster and more stably than Windows XP did.
I would, however, say that I'm guessing that the most earnest Ubuntu users are Linux administrators, who want a nice GUI interface on top of their preferred Linux tools, on the machines they use to connect to the headless servers they support.
I don't think that Linux, as such, will ever quite overcome Microsoft. However, given the overtures towards open source from Microsoft, I think there may be a fruitful convergence.
If we understand improvements in technology as improvements in productivity, then there are a number of ways to split up the benefit from improved productivity: more capital, more leisure time, higher consumption. I think it's not a strange argument to suggest we should re-balance the benefit of increased productivity so that we accumulate capital and increase consumption at slower rates, while gradually increasing leisure time.
I do, however, think that word processors are badly designed from the point of facilitating writing. Most advice on writing encourages the writer to break the process up into separate phases: brainstorming and free-writing, outlining, writing a draft, revising, revising, and revising. Word processors tend to encourage doing all steps at once, and worse, encourage the writer to choose layout and typesetting options before the writer begins writing, when writers generally shouldn't bother about those details at all. Brainstorming and free-writing are widely recommended practices, that most word processors implicitly discourage, with automatic spelling and grammar checking.
Nearly everyone I've known who takes writing seriously, student or professional, struggles with minimizing distractions from the writing process. There's something particularly difficult about writing, the process of putting one's thoughts in words which, in itself, cannot be a clear algorithmic process, and most people will be tempted to procrastinate, in the form of doing something that seems related, but isn't really useful. Word processors, with all their layout tweaks available when clicking on bright, attractive buttons, are full of temptations to procrastinate and distract oneself from the writing itself. Even launching a word processor is significantly slower than launching a text editor, and most include a (distracting) splash screen.
I've never seen a child, assigned to write an essay, who will not fiddle with fonts, layout options, etc., before typing a single word.
Concentrating on writing in a word processor is like meditating in an amusement park -- with sufficient discipline, it can be done, but it's really not a conducive environment.
For writing, I think a better approach is, at least, breaking the software tools into two: the actual writing, and the layout. The latter part could often be optional. Most simple text editors, like Notebook or gedit, are more than adequate for writing, revising, saving, and loading, and include basic spell-checking.
Congressional candidates tend to have strong connections to wealth and political power, and are often lawyers. The DMCA may have pissed off the wrong people, this time.
This comes up a lot, and I've seen similar episodes myself. The question that nags at me is, WHY are people afraid to do something that's easier than toasting bread in a toaster, and far, far easier than driving a car to the store?
Staples Business Depot here sells pre-configured routers at $15 over retail.
That doesn't make any sense. Most routers are pre-configured, as much as they can be, with the factory defaults, and as far as possible, routers configure themselves when plugged in. Further configuration would have to be done with knowledge of the specifics of the network setup, or at least after the router has been connected.
For users who simply want to connect a wi-fi router to their cable modem, there's no pre-configuration that can be done. For something more complicated that the user doesn't know how to handle, or doesn't want to, you'd need to send out a technician, and that would cost more than $15, even at minimum wage.
From what I've read elsewhere,/64 allotments are required for router advertisements in the standard EUI-64 scheme, in which the IPv6 address for a node is composed of the network prefix plus the 64-bit MAC address; so, the maximum length for a network prefix is 64 bits.
By default, the host portion of an IPv6 address is the node's MAC address. Since some are worried that this is a security issue, there's a standard for randomly generating the host portion of an IPv6 address. That's active by default in Windows 7, and easily activated in Linux and OS X. This means that when a Windows 7 box boots up, it will use one IPv6 address temporarily, before generating a new one, and the new one is regenerated, I think, weekly.
So, as things stand now, a single machine is likely to use several different IPv6 addresses in a month, without the user's intervention, on one of the most commonly used operating systems. That makes counting IPs used problematic.
It's not ideal from the customer perspective, certainly. I'm not clear on why it's a dynamic assignment, not static. Still, there are workarounds.
For one thing, dynamically assigned IPv4 addresses are common, and there are ways to cope with the limitations. I've been using Comcast for about a year, and have a dynamically assigned IPv4 address; I think the renewal period is a week. As long as the node keeps renewing the IP address, it should maintain the same one.
Also there's dynamic DNS, in which you report your current IP address to a server, so that the domain name record is updated. IPv6 addresses are longer, and thus even harder to remember and retype accurately, than IPv4 addresses, so I expect name service will become more important than it already is.
One of the great things about IPv6 is router advertisements and auto-configuration, so in most cases, users wouldn't notice any changes when the dynamic assignment changes.
You can assign as many IPv6 addresses to a single node as you like, and there are automatically generated local link addresses that will remain stable, so those would allow for stable IPv6 addresses for use within a local network.
On the whole, though, I think it would be better if ISPs assigned static prefixes. It's not like there will be a desperate need to recycle those prefixes -- they can recycle them when an account is closed, but they really shouldn't need to do it before.
I remember similar policies about ISP contracts only covering one machine, but I would guess they've since been abandoned. Now that you've reminded me, I think NAT was often implicitly marketed to home users as a way to fool ISPs; at some point, it became so common for home users to have multiple nodes that ISPs had to give up the limitation -- customer expectations had shifted too far. Perhaps we could call it "market disobedience," on the model of "civil disobedience."
I would also guess that individually assigning IPv6 addresses would be an administrative nightmare for ISPs, given the scale of the IPv6 namespace, given that everyone who knows what IPv6 is knows that there's practically no limit to the number of available IPv6 addresses, and given that there's reason to expect further growth in the number of Internet-connected devices in homes.
My understanding was that an RFC had recommended that ISPs assign network prefixes of at most/64, and suggested/48 -- if I'm reading it correctly, this is in RFC 4779, section 5.2.2.2.2, but I'm not certain I understand what I'm reading there.
Slashdot Mirror
I didn't know it had to be activated on a computer. I do know there are people who own and use iPhones and do not own personal computers. If there's some initial activation, couldn't they just do that on a computer that they borrow?
How can they use what?
The iPhone itself comes with a bunch of apps pre-installed, including a Web browser. So do other smartphones. I thought that was pretty obvious.
In one of the movies -- the third one, I think -- Magneto gives a moving speech, the narrative purpose of which I believe was to shake up the audience's assumptions about which side had just the greater claim to justice.
Then, Magneto ordered one of his followers to kill an innocent person.
I've seen this sort of thing elsewhere, and I wonder if there's a term for it on TVTropes: when there's a strong suggestion that the villain isn't really a villain after all, and then this suggestion is immediately subverted by the villain doing something horrible without any justification.
Come to think of it, I remember comparing notes with other people about the first movie, and nearly everyone agreed that Magneto's plan to turn anti-mutant political leaders into mutants actually seemed like a good idea, but it was also subverted by his kidnapping and torture of Rogue.
On reflection, I understated the significance of ATMs using two-factor authentication. That it's two distinct tests probably matters much more than how strong a password a PIN is.
I'm beginning to think there's excessive paranoia about a very narrow conception of security.
I am asked to invent at least one password a day. Most often, it's for something for which there is no need for any security. In fact, given that users, who are asked to make up passwords frequently, at a moment's notice, understandably reuse passwords, requiring a password where it isn't really necessary actually undermines security.
Every bank I've used, and as far as I know, every bank, requires a four-digit PIN for ATM access. That means that the single most important password most people use daily is a very weak one -- offset by the second form of authentication, the ATM card. ATM cards are frequently left in ATM machines by mistake. How much time do people spend worrying about the security of their bank PINs?
There's an issue going unaddressed here: the idea of a "superhero" is based upon the idea of a "hero," and that is an explicitly aristocratic and anti-egalitarian idea.
The classic hero is a legendary figure who is also supposed to be a group's ancestor, and aristocrats are traditionally direct descendants of heroes. Take the heroes in the Iliad and Odyssey, for example: they are explicitly aristocrats, the ancestors of aristocratic lineages, and they are specially connected to gods and have abilities that set them beyond ordinary people.
The egalitarian ideal posits that people don't vary that much in ability. Not, of course, that everyone is actually equal in ability, but that people don't vary so much that one person should be absolutely elevated over another. When one person has significantly more power than another, it's a morally perilous situation; the person with more power is expected to express humility and restraint.
For every Einstein, there are hundreds of thousands of graduate students who understand Einstein's theories, millions of college students who understand them in outline, and countless high school students who have a cursory understanding of energy-matter equivalence and the twin paradox. Shakespeare had many contemporaries, scarcely known now except by students of drama and literary history, who wrote plays with similar themes and styles. While there are differences in abilities, we tend to oversimplify history and exaggerate those differences.
The egalitarian ideal is based upon the defeat of the aristocratic ideal -- superhero stories, like a lot of fantasy fiction, include aristocratic ideals trying to slip back in.
Stories about "superheroes" have premises that make for ongoing internal and external conflicts: how can you reconcile egalitarian ideals with the existence of individuals with extraordinary power? Thus, the general insistence on secret identities, Superman's discipline of self-restraint, Spiderman's ethical agonizing, the X-Men's struggles with democracy and its subversion. What doesn't seem to be acknowledged is that the egalitarian ideal has a fundamental premise that (super)heroes don't really exist. Consequently, taking the superhero stories literally, the struggles of "good" superheroes to maintain ideals that are premised upon their non-existence fundamentally don't make sense. They only make sense as storytelling because the "good" superheroes keep struggling to support our reality, not their own.
So, yes, if someone is asked to imagine being a superhero, they're imagining themselves as figures that are presumed not to exist, in a real world context -- they're imagining themselves as beings that, by existing, break established social rules. Thus, the choice is between villainy, and a hidden aristocracy -- the latter not actually making much sense. But, imagining a world in which superheroes actually exist, in a self-consistent way, means imagining a world in which aristocratic ideals are valid -- which immediately strikes most people as loathsome.
I'm a little chagrined, as I should have been conscious of the distinctions.
Anyway, I checked the man page and tested the options. "sudo -s" will inherit the environment from the user who executes it -- i.e., use your path, stay in the directory where you executed "sudo -s", etc. So would "sudo bash".
"sudo -i" will simulate logging in as another user -- by default, root -- and so instead of inheriting the environment, will execute the login configuration files for that user.
So, "sudo -i" is a better equivalent to "su -", and "sudo -s" would be the one to use if you want to use the same environment as the regular user.
It's well known that for message boards, blogs, email lists, etc., most readers are "lurkers," who read without posting responses. I expect the metrics vary, but I remember someone's general rule was one hundred lurkers for every active participant in a discussion.
So, I read these stats as meaning that there's an astonishingly high level of response on Twitter -- although that's offset from the comments that there are people gaming Twitter to generate "hype."
Granted, there's the introductory video from Twitter that suggests you use Twitter in just that way.
However, almost everyone I know who uses Twitter uses it primarily to post links to articles or interesting Websites, along with a short comment on the content -- hence the wide use of URL shorteners.
You can enable the root account in Ubuntu if you really want, but it's recommended against. For that matter, logging straight in as root on any *nix system is generally considered a bad idea, security-wise -- standard practice on other systems is to log in as a normal user and use "su -" to switch to a root shell. You definitely shouldn't log straight in as root in a GUI.
Ubuntu (like OS X) uses sudo, and in practice, in the GUI, it means that a user that is part of the admin group is prompted for a password when executing an admin function, like installing a package system-wide, which is a pretty reasonable lower limit for security. At the command line, if you want a root shell, you could just do "sudo -s", which is effectively the same as "su -" on other *nix systems.
Oh, and if you want to run Java plugins, just install "icedtea6-plugin" from Synaptic, or from the command line with "sudo apt-get install icedtea6-plugin".
It's been in the Ubuntu repositories since at least 8.04. It used to be a bit buggy, but it got markedly more reliable a few distributions ago: 9.04 or 9.10, I think.
Every time I ride the bus, I am struck by how many people are using iPhones -- particularly the poor people. It's counter-intuitive, until you realize that if you can only afford one Internet-capable device, instead of several, you're going to choose the one that does the most, which is most likely an iPhone.
I think a more accurate picture is that Microsoft is challenged on the enterprise end by Linux, on the consumer end by Apple, and in the Internet by Google. Microsoft is powerful, but boxed in. I think there are signs that Microsoft is exploring alliances with the open source community in order to break out of that box.
While it's easier to network Windows desktops with Linux servers than it is to network Linux desktops with Windows servers, it's still easier to network Linux desktops with Linux servers. IT staffs tend to prefer homogeneous environments. If Linux servers continue to be preferred, there will be some pressure to shift from Windows desktops to Linux desktops -- which is where the end-of-life of Windows XP is relevant.
Many people are unsatisfied with OpenOffice, but I don't hear a lot of praise for the most recent version of Microsoft Office, either, and that's the killer app for Windows in the workplace. At some point, if OpenOffice or some other office suite breaks through into general use, there will be a much stronger challenge to Window's dominance on the desktop.
I'm using a computer that's about six years old, and I dual-boot Ubuntu 10.10 and Windows 7. Both work well. On Windows, I have trouble with newer games, but that's to be expected; I find Windows 7 runs faster and more stably than Windows XP did.
I wouldn't go so far as to say the dream is dead.
I would, however, say that I'm guessing that the most earnest Ubuntu users are Linux administrators, who want a nice GUI interface on top of their preferred Linux tools, on the machines they use to connect to the headless servers they support.
I don't think that Linux, as such, will ever quite overcome Microsoft. However, given the overtures towards open source from Microsoft, I think there may be a fruitful convergence.
That's true if you take it to that extreme.
If we understand improvements in technology as improvements in productivity, then there are a number of ways to split up the benefit from improved productivity: more capital, more leisure time, higher consumption. I think it's not a strange argument to suggest we should re-balance the benefit of increased productivity so that we accumulate capital and increase consumption at slower rates, while gradually increasing leisure time.
I believe that's the primary point.
I do, however, think that word processors are badly designed from the point of facilitating writing. Most advice on writing encourages the writer to break the process up into separate phases: brainstorming and free-writing, outlining, writing a draft, revising, revising, and revising. Word processors tend to encourage doing all steps at once, and worse, encourage the writer to choose layout and typesetting options before the writer begins writing, when writers generally shouldn't bother about those details at all. Brainstorming and free-writing are widely recommended practices, that most word processors implicitly discourage, with automatic spelling and grammar checking.
Nearly everyone I've known who takes writing seriously, student or professional, struggles with minimizing distractions from the writing process. There's something particularly difficult about writing, the process of putting one's thoughts in words which, in itself, cannot be a clear algorithmic process, and most people will be tempted to procrastinate, in the form of doing something that seems related, but isn't really useful. Word processors, with all their layout tweaks available when clicking on bright, attractive buttons, are full of temptations to procrastinate and distract oneself from the writing itself. Even launching a word processor is significantly slower than launching a text editor, and most include a (distracting) splash screen.
I've never seen a child, assigned to write an essay, who will not fiddle with fonts, layout options, etc., before typing a single word.
Concentrating on writing in a word processor is like meditating in an amusement park -- with sufficient discipline, it can be done, but it's really not a conducive environment.
For writing, I think a better approach is, at least, breaking the software tools into two: the actual writing, and the layout. The latter part could often be optional. Most simple text editors, like Notebook or gedit, are more than adequate for writing, revising, saving, and loading, and include basic spell-checking.
Don't be silly. "tl;dr" is an Internet acronym for, "I'm stupid and lazy."
Congressional candidates tend to have strong connections to wealth and political power, and are often lawyers. The DMCA may have pissed off the wrong people, this time.
This comes up a lot, and I've seen similar episodes myself. The question that nags at me is, WHY are people afraid to do something that's easier than toasting bread in a toaster, and far, far easier than driving a car to the store?
Why is it that people are afraid of computers?
Staples Business Depot here sells pre-configured routers at $15 over retail.
That doesn't make any sense. Most routers are pre-configured, as much as they can be, with the factory defaults, and as far as possible, routers configure themselves when plugged in. Further configuration would have to be done with knowledge of the specifics of the network setup, or at least after the router has been connected.
For users who simply want to connect a wi-fi router to their cable modem, there's no pre-configuration that can be done. For something more complicated that the user doesn't know how to handle, or doesn't want to, you'd need to send out a technician, and that would cost more than $15, even at minimum wage.
From what I've read elsewhere, /64 allotments are required for router advertisements in the standard EUI-64 scheme, in which the IPv6 address for a node is composed of the network prefix plus the 64-bit MAC address; so, the maximum length for a network prefix is 64 bits.
By default, the host portion of an IPv6 address is the node's MAC address. Since some are worried that this is a security issue, there's a standard for randomly generating the host portion of an IPv6 address. That's active by default in Windows 7, and easily activated in Linux and OS X. This means that when a Windows 7 box boots up, it will use one IPv6 address temporarily, before generating a new one, and the new one is regenerated, I think, weekly.
So, as things stand now, a single machine is likely to use several different IPv6 addresses in a month, without the user's intervention, on one of the most commonly used operating systems. That makes counting IPs used problematic.
It's not ideal from the customer perspective, certainly. I'm not clear on why it's a dynamic assignment, not static. Still, there are workarounds.
For one thing, dynamically assigned IPv4 addresses are common, and there are ways to cope with the limitations. I've been using Comcast for about a year, and have a dynamically assigned IPv4 address; I think the renewal period is a week. As long as the node keeps renewing the IP address, it should maintain the same one.
Also there's dynamic DNS, in which you report your current IP address to a server, so that the domain name record is updated. IPv6 addresses are longer, and thus even harder to remember and retype accurately, than IPv4 addresses, so I expect name service will become more important than it already is.
One of the great things about IPv6 is router advertisements and auto-configuration, so in most cases, users wouldn't notice any changes when the dynamic assignment changes.
You can assign as many IPv6 addresses to a single node as you like, and there are automatically generated local link addresses that will remain stable, so those would allow for stable IPv6 addresses for use within a local network.
On the whole, though, I think it would be better if ISPs assigned static prefixes. It's not like there will be a desperate need to recycle those prefixes -- they can recycle them when an account is closed, but they really shouldn't need to do it before.
I remember similar policies about ISP contracts only covering one machine, but I would guess they've since been abandoned. Now that you've reminded me, I think NAT was often implicitly marketed to home users as a way to fool ISPs; at some point, it became so common for home users to have multiple nodes that ISPs had to give up the limitation -- customer expectations had shifted too far. Perhaps we could call it "market disobedience," on the model of "civil disobedience."
I would also guess that individually assigning IPv6 addresses would be an administrative nightmare for ISPs, given the scale of the IPv6 namespace, given that everyone who knows what IPv6 is knows that there's practically no limit to the number of available IPv6 addresses, and given that there's reason to expect further growth in the number of Internet-connected devices in homes.
My understanding was that an RFC had recommended that ISPs assign network prefixes of at most /64, and suggested /48 -- if I'm reading it correctly, this is in RFC 4779, section 5.2.2.2.2, but I'm not certain I understand what I'm reading there.