First off it sounds like you can't have a central Java policy but rather need multiple Java's that are department specific.
Oracle needs to figure out how to do security patches that just fix the vulnerabilities and don't introduce (and remove) features.
There are two things here:
1) Bug for bug compatible. Which is something that enterprises want but in general most software vendors don't want to provide. 2) Minor vs. major vs. bug fix version numbers, which I wish people would go back to.
Why the hell don't we use RSS (or equivalent) for this yet, and be able to group all the updates together in a single interface, with a single "update now" button?
Remember to make this comment next time you hear people complaining about the Linux distribution centered model for software. In any case Apple is moving towards this with the App store.
Because users are not the target audience for this lobbying. It is distribution makers and programmers. Things like the free lists from Debian have for almost 20 years been extremely important for what gets casually bundled with other programs. If App X crucially depends on Y which is non free then X is considered non free. But if there is a free alternative to Y, like Z then X's author can often make the dependency Z not Y so that he can be free. X's author may not have been aware of Y's status but getting cut from lists like Debian free makes them aware of it. This mechanism prevents chains of non free licenses from developing which make the whole OS essentially unfree.
Look at almost any major package in Linux that's been around for 15 years and you'll see a battle over this sort of casual chain where the punishment of being considered nonfree has been enough to get this isolation. Big vendors: Sun, Apple, Oracle, IBM, SGI, HP, Nokia have all gotten stung and then come into compliance. It is a true success story.
Let me just throw in one more comment as I read a bit of your discussion. Raenex is claiming that the FSF's understanding is irrelevant. I made allusion to this in my other post but let me make it explicit. The FSF wrote the license. Everyone agrees that G in GPL stands for GNU (i.e. the project of the FSF). The FSF's published interpretation of the license is absolutely relevant to what the license means. The idea that the understand of the parties to a contract governs the interpretation of a contract is part of the very basic of contract law.
In particular copyright law, these documents are part of what is called the "form" of the license. They aren't 100% determinative but they have real and substantial weight.
Tepples I think you are missing something really key here...
Consider a typical album being sold.
a) The producer gets a license for each and every song on the album. That is he has license or copyright for the songs themselves.
b) He then releases the album. He has copyright for the album.
There is very little law governing software. That model I just gave you is the way the courts choose to look at software. If you want to argue about whether software should be treated like a record or should be treated like a book full of articles, that's a valid argument about legal theory. But if you want to talk about the law as it actually exists, the black letter law governing music conglomerations is the law underying the common law which is led to court rulings governing software. But lets not confuse the debate about what should apply and what does apply, i.e. what you think the law should be with what it is.
So now with that in ming here is why I believe the law is correct in treating software like music and not literature. The core reason is because music like software is frequently created on a work for service model i.e. there musicians / programmers who are paid employees and will surrender all copyright for their creative production. Second, there are masters (which have copyright) which go through an editing processes from which record version is derived which has another copyright. This is analogous to compiling source and distributing binaries. Literature does not go this sort of process, the typesetting process, except for very heavy layout materials like magazines is not substantial enough to warrant this extra layer of protection.
Now with that in mind.... A distribution is an album with a bunch of songs (programs) on it. There is a copyright for the distribution, which Ubuntu for example holds, and Ubuntu is claiming to have a license for all other programs/songs on this album. They might be right or wrong in this claim. If someone who wrote program (A) and granted their license to redistribute to Ubuntu wants to claim that including program (B) violates the license for (A), they are going to need to show very strong evidence in the license. There is no question where the burdon is going to lie and it is going to be very very high.
Given that GPLv2 program is and has been distributed with program (B) or programs like B for years, with knowledge and yet no apparent objection, it doesn't even matter anymore what the license says (A) would lose. (A) could start asserting their rights in a more explicit way but that would changing the license for (A) as a first step. The law is not going to permit a copyright holder who writes a vague ambiguous license to assert a serious violation has occurred. What the courts are going to say is essentially "make the license stop being vague on this point or STFU". The fact that there is widespread disagreement is enough to prove that the license isn't unambiguous enough to find that Ubuntu has committed a tort.
Extremely importantly, Debian, which has association with the FSF also does this. Which means the form of the license on A does allow it, which is another barrier.
Now the Linux kernel as our (A) is a particularly bad case because AFAIKT no one has enough standing to assert a license violation of the kernel but only small parts of the kernel. I.E. the kernel itself is an aggregation where the kernel team is distributing a record.
Now if you and and Raenex about what the license should mean under a hypothetical of an entirely different body of law fine. But if you want to know why his argument is wrong, it is wrong because lots of people think it is wrong and that's all that really matters.
I get that. I honestly don't see nothing but improvement between 10.0 and 10.6 in terms of functionality. The plumbing has gotten much better and that has led to all sorts of interface enhancements in areas like drag and drop and selection that I use daily.
I like the direction 10.7 is going in moving applications away from hierarchical filesystem type approach which originated out of dual floppies and towards a mini computer like database filesystem. The all applications are always running and save state of 10.7 is a huge step up. To accomplish this they've made spaces a bit easier for general users which is a bit of a downgrade. But as someone who was using virtual desktops on OSX since 10.1 getting applications to be designed in a virtual desktop friendly way is a huge plus even if the virtual desktop management (mission control) has gotten a bit more naive than 10.6's spaces.
No question Linux has vastly more sophisticated window managers like XMonad or Enlightenment than either Windows or OSX, it isn't even close. If you want sophistication it is a no brainer... but in practice even Linux users seem to want a simple reliable window manager and not a feature rich window manager. Qnome, KDE, LXDE... all use simple window managers. ELive which is the distribution that most focused on enlightenment features:
a) aimed for legacy hardware i.e. a way to make your old system cool, not a primary system b) isn't very popular and does a major release about every 3 years.
Quartz-wm while a reliable X11 window manager, mainly just offers excellent integration with Aqua. I don't really care that much about my window manger. But if you think you might want the power I'd say if you are using OSX start X11 in full screen mode and play with either one of those two and see if it is worth it.
I'd love to see Apple have a tiling mode for Quartz.
Oh I agree on the value add its huge. You can get that with Linux by buying a Linux laptop. Or the OEMs are pretty good on doing that sort of stuff with server class hardware. The problem is that Linux has done a great job on securing the cheap bottom part of the market that doesn't want to pay for the value add.
OTOH I'm mainly an OSX user. Linux Unix applications are often better configured than their MacPorts or Fink variants. So for complex Unix install the time spent configuring drivers is way overshadowed by the time spent configuring complex application. But most of those apps run on developer workstations or servers.
I'm pretty much of the opinion that if you don't have a good reason to want a Unix, Linux still doesn't make sense. And I say that as a guy who has been a regular Linux user since '95 and a regular Unix user since '88.
Well that's a flaw right there. No that's now the definition The material object, other than a phonorecord, in which the copyrighted work is first fixed, and from which the work can be perceived, reproduced, or otherwise communicated, either directly or with the aid of a machine or device.
For computers the "phonorecord" standard applies i.e. a CD is automatically an aggregation. Copyright in a sound recording protects the particular series of sounds embodied in the sound recording. Copyright registration for a sound recording alone is not the same as registration for the musical, dramatic, or literary work recorded. The underlying work may be registered in its own right apart from any recording of the performance.
I.E. (we think) software is copyrighted individually and the complication / aggregation is also copyrighted and those are two different copyrights.
____
As for the bandwidth not being available I've lost you. If you don't use computers and just use phones then this conversation doesn't apply. If you use computers then mostly to use the internet on them (wifi or ethernet cable) you have a land connection based connection and aren't paying $10 GB. If you don't have a land based connection then go back to sneakernet, use the mail http://www.osdisc.com/
I think your comments regarding iOS like features are a bit overblown. Nothing is changing too much for "real work". But if you want a Unixy GUI under OSX, OSX ships with X.app and you can run a different GUI. Just go to http://www.macports.org/ or http://www.finkproject.org/ and install the GUI you want.
I still don't get why the other boxes it worked fine with same hardware but that is the quirks...Crap happens but in my experience hardware related crap happens more often with Linux, sound not working, wifi, graphics etc.
That's what a multi billion dollar test lab like Microsoft buys you, along with millions of special cases interlaced all over the software. Lets say for example the hardware vendor had a soldering problem so something the hardware bus doesn't boot properly. The Windows kernel when it detects the machine uses a longer delay and boots the bus twice. Linux is awesome when you get that kind of problem from breakage because you can do that sort of stuff yourself.
It's not exactly free if one has to pay the telco $7 (at $10/GB) for the 700 MB overage.
Why are you downloading data to burn using over-the-air. Over-the-air is for travel. That's like saying it isn't exactly free if you have to hire a limousine to drive down to Staples.
Not if someone tries to assert Raenex's interpretation of GPLv2 against Canonical and other distributors of Linux operating systems that include non-free drivers.
The problem with Raenex's interpretation is that Canonical still works fully on some hardware with only the free drivers. So I think you can see this as an aggregation of multiple works. That being said, we could really use much more specific black letter copyright law with regard to software so that everyone can move from pretty good reasons why stuff is probably true to explicit citations.
You are thinking it was one person. Most likely you had different people controlling different parts of the project with different agendas. The asphalt guy may very well have wanted that tree out of their and lost the vote in committee.
If GNU/Linux wants wider adoption, hardware compatibility and the concept of "it just works" is still a major problem.
Microsoft spends billions per year every year on making sure their OS works fantastically with the incredible range of semi-compliant crappy hardware that is in the consumer space. And that BTW is with OEM's considering Windows support to be a top priority. Asking Linux to have that level of support is impossible, it can't be done. You were accusing the parent of living in a fantasy world, expecting Linux to easily and fully support that range of hardware without any vendor support for around $20m global is a fantasy as well.
Linux has managed to have a better range of hardware support than any other OS besides Windows. I'll agree there is a huge range between the #1 slot and the #2 slot.
I like that the Linux community aimed to be more than just a free Unix and actually become an alternative for Windows power users. But I think ultimately it is a bridge too far. Linux is far and away the easiest Unix to install, if you for some reason are highly motivated to install a Unix which is different from saying it is or will be as easy as Windows.
The point of those distributions is not high usage or really even any usage but to prevent a gradually slipping towards a world where free and non-free are intermixed so casually that Linux becomes unfree. Debian does a good job of that, but they do a good job of that because they have a substantial number of people who care about these issues.
Ubuntu offers an OEM specific hardware specific repository / version. The OEM would just have the default point to the limited repository and not the main one. A version upgrade would be a choice for the OEM of whether they wanted to pay for a new version. So Ubuntu already and has for several years offered the service you are asking for.
Since we are talking the Asian market there are several good Linux vendors like Linpus where are tightly focused on Linuxes that work on specific hardware. Their system, because there is no general repository, doesn't allow the end user to bork their machine unless they wipe the Linux entirely.
If you are talking Canada I'm not even sure what the issues are regarding bandwidth or carriers or subsidies or.... Everything I wrote was very US specific.
As for carriers selling wifi, if the cost of their bandwidth is high enough in cities it might make sense. They could do something like 10m wifi = 1m over-air so they can still charge yet still encourage wifi.
Oh all things being equal the carriers love Android and hate Apple. But things aren't equal. The 30% is just a tip but the extra $50 in subsidy really bites. Worse than that though is the brand loyalty issue. If Verizon stopped carrying Motorola no one is following them out the door. But RIM used to and Apple does have the kind of brand loyalty where customers would change carriers to get the phone in large numbers. And those are the least price sensitive, i.e. highest margin customers of Verizon.
If you come in looking to buy an iPhone they aren't going to be a problem. If you are iffy they will quite often steer you towards Android.
I forgot to mention I've been a Verizon customer since about 2000. They tried to steer me away from iPhone when they first got it, and I was thinking of paying the penalty to get a phone sooner.
The customers you are talking about the loyal to semi-loyal Apple people are Verizon's high margin customers. They don't want to steer them anywhere but towards more accessories and additional services. But there is a huge group of consumers who know about the iPhone but are very iffy about it. On the surface, and quite often in reality the Android phones can look like a better value for what you get: even though the iPhone subsidies are higher Apple's profit margins can be much higher. Verizon also has pretty good loyalty because of quality of service.
In any case this sort of steering isn't going to cost them much. Apple likes being with the premium carrier and as long as they get subsidies they aren't leaving. Verizon likes having the premium phone and the high margin customer base. It is a good partnership.
suppose every single oil company in the world is doing an absolutely amazing job as we all pay $4/gallon for gas, right?
The oil companies don't set gas prices, that's the commodities market. As for doing a great job in the last 5 years they've boosted production from 85m barrels / day to 90m against a world where easy to get to reserves are collapsing. Yes they are doing a great job. They are making far too much money for that great job, but gas would be $75/gallon if it wasn't for the technologies they introduced in the last generation.
If that were true, then care to explain how we came to define and follow Moores law for many a decade?
What does the transistor density have to do with signal error rates?
Sorry, but an unlimited data plan is not some sort of physical impossibility.
Actually there are limits. That was his point. As for unlimited internet in other markets, generally those are capped to. But over the air is vastly more complex than wired so you bump against the one limit and not the other. People who backup their entire Blu-Ray collection notice the home internet limits.
If all phones would default to using wifi for everything whenever it is available then it would take a huge chunk of the load off of the cell towers.
The iPhone does that. I don't know that the carriers want to force this though. And no this is not artificial. The real problem is congress not cannibalizing the HDTV bandwidth for over the air internet. Who uses rabbit ears anymore?
Slashdot Mirror
First off it sounds like you can't have a central Java policy but rather need multiple Java's that are department specific.
Oracle needs to figure out how to do security patches that just fix the vulnerabilities and don't introduce (and remove) features.
There are two things here:
1) Bug for bug compatible. Which is something that enterprises want but in general most software vendors don't want to provide.
2) Minor vs. major vs. bug fix version numbers, which I wish people would go back to.
Why the hell don't we use RSS (or equivalent) for this yet, and be able to group all the updates together in a single interface, with a single "update now" button?
Remember to make this comment next time you hear people complaining about the Linux distribution centered model for software. In any case Apple is moving towards this with the App store.
Because users are not the target audience for this lobbying. It is distribution makers and programmers. Things like the free lists from Debian have for almost 20 years been extremely important for what gets casually bundled with other programs. If App X crucially depends on Y which is non free then X is considered non free. But if there is a free alternative to Y, like Z then X's author can often make the dependency Z not Y so that he can be free. X's author may not have been aware of Y's status but getting cut from lists like Debian free makes them aware of it. This mechanism prevents chains of non free licenses from developing which make the whole OS essentially unfree.
Look at almost any major package in Linux that's been around for 15 years and you'll see a battle over this sort of casual chain where the punishment of being considered nonfree has been enough to get this isolation. Big vendors: Sun, Apple, Oracle, IBM, SGI, HP, Nokia have all gotten stung and then come into compliance. It is a true success story.
Let me just throw in one more comment as I read a bit of your discussion. Raenex is claiming that the FSF's understanding is irrelevant. I made allusion to this in my other post but let me make it explicit. The FSF wrote the license. Everyone agrees that G in GPL stands for GNU (i.e. the project of the FSF). The FSF's published interpretation of the license is absolutely relevant to what the license means. The idea that the understand of the parties to a contract governs the interpretation of a contract is part of the very basic of contract law.
In particular copyright law, these documents are part of what is called the "form" of the license. They aren't 100% determinative but they have real and substantial weight.
Tepples I think you are missing something really key here...
Consider a typical album being sold.
a) The producer gets a license for each and every song on the album. That is he has license or copyright for the songs themselves.
b) He then releases the album. He has copyright for the album.
There is very little law governing software. That model I just gave you is the way the courts choose to look at software. If you want to argue about whether software should be treated like a record or should be treated like a book full of articles, that's a valid argument about legal theory. But if you want to talk about the law as it actually exists, the black letter law governing music conglomerations is the law underying the common law which is led to court rulings governing software. But lets not confuse the debate about what should apply and what does apply, i.e. what you think the law should be with what it is.
So now with that in ming here is why I believe the law is correct in treating software like music and not literature. The core reason is because music like software is frequently created on a work for service model i.e. there musicians / programmers who are paid employees and will surrender all copyright for their creative production. Second, there are masters (which have copyright) which go through an editing processes from which record version is derived which has another copyright. This is analogous to compiling source and distributing binaries. Literature does not go this sort of process, the typesetting process, except for very heavy layout materials like magazines is not substantial enough to warrant this extra layer of protection.
Now with that in mind.... A distribution is an album with a bunch of songs (programs) on it. There is a copyright for the distribution, which Ubuntu for example holds, and Ubuntu is claiming to have a license for all other programs/songs on this album. They might be right or wrong in this claim. If someone who wrote program (A) and granted their license to redistribute to Ubuntu wants to claim that including program (B) violates the license for (A), they are going to need to show very strong evidence in the license. There is no question where the burdon is going to lie and it is going to be very very high.
Given that GPLv2 program is and has been distributed with program (B) or programs like B for years, with knowledge and yet no apparent objection, it doesn't even matter anymore what the license says (A) would lose. (A) could start asserting their rights in a more explicit way but that would changing the license for (A) as a first step. The law is not going to permit a copyright holder who writes a vague ambiguous license to assert a serious violation has occurred. What the courts are going to say is essentially "make the license stop being vague on this point or STFU". The fact that there is widespread disagreement is enough to prove that the license isn't unambiguous enough to find that Ubuntu has committed a tort.
Extremely importantly, Debian, which has association with the FSF also does this. Which means the form of the license on A does allow it, which is another barrier.
Now the Linux kernel as our (A) is a particularly bad case because AFAIKT no one has enough standing to assert a license violation of the kernel but only small parts of the kernel. I.E. the kernel itself is an aggregation where the kernel team is distributing a record.
Now if you and and Raenex about what the license should mean under a hypothetical of an entirely different body of law fine. But if you want to know why his argument is wrong, it is wrong because lots of people think it is wrong and that's all that really matters.
I get that. I honestly don't see nothing but improvement between 10.0 and 10.6 in terms of functionality. The plumbing has gotten much better and that has led to all sorts of interface enhancements in areas like drag and drop and selection that I use daily.
I like the direction 10.7 is going in moving applications away from hierarchical filesystem type approach which originated out of dual floppies and towards a mini computer like database filesystem. The all applications are always running and save state of 10.7 is a huge step up. To accomplish this they've made spaces a bit easier for general users which is a bit of a downgrade. But as someone who was using virtual desktops on OSX since 10.1 getting applications to be designed in a virtual desktop friendly way is a huge plus even if the virtual desktop management (mission control) has gotten a bit more naive than 10.6's spaces.
No question Linux has vastly more sophisticated window managers like XMonad or Enlightenment than either Windows or OSX, it isn't even close. If you want sophistication it is a no brainer... but in practice even Linux users seem to want a simple reliable window manager and not a feature rich window manager. Qnome, KDE, LXDE... all use simple window managers. ELive which is the distribution that most focused on enlightenment features:
a) aimed for legacy hardware i.e. a way to make your old system cool, not a primary system
b) isn't very popular and does a major release about every 3 years.
Quartz-wm while a reliable X11 window manager, mainly just offers excellent integration with Aqua. I don't really care that much about my window manger. But if you think you might want the power I'd say if you are using OSX start X11 in full screen mode and play with either one of those two and see if it is worth it.
I'd love to see Apple have a tiling mode for Quartz.
Oh I agree on the value add its huge. You can get that with Linux by buying a Linux laptop. Or the OEMs are pretty good on doing that sort of stuff with server class hardware. The problem is that Linux has done a great job on securing the cheap bottom part of the market that doesn't want to pay for the value add.
OTOH I'm mainly an OSX user. Linux Unix applications are often better configured than their MacPorts or Fink variants. So for complex Unix install the time spent configuring drivers is way overshadowed by the time spent configuring complex application. But most of those apps run on developer workstations or servers.
I'm pretty much of the opinion that if you don't have a good reason to want a Unix, Linux still doesn't make sense. And I say that as a guy who has been a regular Linux user since '95 and a regular Unix user since '88.
The law defines a "copy" as one physical medium
Well that's a flaw right there. No that's now the definition The material object, other than a phonorecord, in which the copyrighted work is first fixed, and from which the work can be perceived, reproduced, or otherwise communicated, either directly or with the aid of a machine or device.
For computers the "phonorecord" standard applies i.e. a CD is automatically an aggregation. Copyright in a sound recording protects the particular series of sounds embodied in the sound recording. Copyright registration for a sound recording alone is not the same as registration for the musical, dramatic, or literary work recorded. The underlying work may be registered in its own right apart from any recording of the performance.
I.E. (we think) software is copyrighted individually and the complication / aggregation is also copyrighted and those are two different copyrights.
____
As for the bandwidth not being available I've lost you. If you don't use computers and just use phones then this conversation doesn't apply. If you use computers then mostly to use the internet on them (wifi or ethernet cable) you have a land connection based connection and aren't paying $10 GB. If you don't have a land based connection then go back to sneakernet, use the mail http://www.osdisc.com/
If you want to use Linux fine.
I think your comments regarding iOS like features are a bit overblown. Nothing is changing too much for "real work". But if you want a Unixy GUI under OSX, OSX ships with X.app and you can run a different GUI. Just go to http://www.macports.org/ or http://www.finkproject.org/ and install the GUI you want.
I still don't get why the other boxes it worked fine with same hardware but that is the quirks...Crap happens but in my experience hardware related crap happens more often with Linux, sound not working, wifi, graphics etc.
That's what a multi billion dollar test lab like Microsoft buys you, along with millions of special cases interlaced all over the software. Lets say for example the hardware vendor had a soldering problem so something the hardware bus doesn't boot properly. The Windows kernel when it detects the machine uses a longer delay and boots the bus twice. Linux is awesome when you get that kind of problem from breakage because you can do that sort of stuff yourself.
It's not exactly free if one has to pay the telco $7 (at $10/GB) for the 700 MB overage.
Why are you downloading data to burn using over-the-air. Over-the-air is for travel. That's like saying it isn't exactly free if you have to hire a limousine to drive down to Staples.
Not if someone tries to assert Raenex's interpretation of GPLv2 against Canonical and other distributors of Linux operating systems that include non-free drivers.
The problem with Raenex's interpretation is that Canonical still works fully on some hardware with only the free drivers. So I think you can see this as an aggregation of multiple works. That being said, we could really use much more specific black letter copyright law with regard to software so that everyone can move from pretty good reasons why stuff is probably true to explicit citations.
You are thinking it was one person. Most likely you had different people controlling different parts of the project with different agendas. The asphalt guy may very well have wanted that tree out of their and lost the vote in committee.
If GNU/Linux wants wider adoption, hardware compatibility and the concept of "it just works" is still a major problem.
Microsoft spends billions per year every year on making sure their OS works fantastically with the incredible range of semi-compliant crappy hardware that is in the consumer space. And that BTW is with OEM's considering Windows support to be a top priority. Asking Linux to have that level of support is impossible, it can't be done. You were accusing the parent of living in a fantasy world, expecting Linux to easily and fully support that range of hardware without any vendor support for around $20m global is a fantasy as well.
Linux has managed to have a better range of hardware support than any other OS besides Windows. I'll agree there is a huge range between the #1 slot and the #2 slot.
I like that the Linux community aimed to be more than just a free Unix and actually become an alternative for Windows power users. But I think ultimately it is a bridge too far. Linux is far and away the easiest Unix to install, if you for some reason are highly motivated to install a Unix which is different from saying it is or will be as easy as Windows.
The point of those distributions is not high usage or really even any usage but to prevent a gradually slipping towards a world where free and non-free are intermixed so casually that Linux becomes unfree. Debian does a good job of that, but they do a good job of that because they have a substantial number of people who care about these issues.
Ubuntu offers an OEM specific hardware specific repository / version. The OEM would just have the default point to the limited repository and not the main one. A version upgrade would be a choice for the OEM of whether they wanted to pay for a new version. So Ubuntu already and has for several years offered the service you are asking for.
Since we are talking the Asian market there are several good Linux vendors like Linpus where are tightly focused on Linuxes that work on specific hardware. Their system, because there is no general repository, doesn't allow the end user to bork their machine unless they wipe the Linux entirely.
Many of the home connections have monthly limits or throttle. For example comcast in many locations has a 250g / mo limit on their home connection.
Those were 2 different attacks by 2 different groups. The regulators were here. We've also seen the chinese test some capabilities.
Maybe fair would be you get attack for:
a) Market share
b) Corporate espionage
c) Military
If you are talking Canada I'm not even sure what the issues are regarding bandwidth or carriers or subsidies or.... Everything I wrote was very US specific.
As for carriers selling wifi, if the cost of their bandwidth is high enough in cities it might make sense. They could do something like 10m wifi = 1m over-air so they can still charge yet still encourage wifi.
Oh all things being equal the carriers love Android and hate Apple. But things aren't equal. The 30% is just a tip but the extra $50 in subsidy really bites. Worse than that though is the brand loyalty issue. If Verizon stopped carrying Motorola no one is following them out the door. But RIM used to and Apple does have the kind of brand loyalty where customers would change carriers to get the phone in large numbers. And those are the least price sensitive, i.e. highest margin customers of Verizon.
If you come in looking to buy an iPhone they aren't going to be a problem. If you are iffy they will quite often steer you towards Android.
I forgot to mention I've been a Verizon customer since about 2000. They tried to steer me away from iPhone when they first got it, and I was thinking of paying the penalty to get a phone sooner.
The customers you are talking about the loyal to semi-loyal Apple people are Verizon's high margin customers. They don't want to steer them anywhere but towards more accessories and additional services. But there is a huge group of consumers who know about the iPhone but are very iffy about it. On the surface, and quite often in reality the Android phones can look like a better value for what you get: even though the iPhone subsidies are higher Apple's profit margins can be much higher. Verizon also has pretty good loyalty because of quality of service.
In any case this sort of steering isn't going to cost them much. Apple likes being with the premium carrier and as long as they get subsidies they aren't leaving. Verizon likes having the premium phone and the high margin customer base. It is a good partnership.
suppose every single oil company in the world is doing an absolutely amazing job as we all pay $4/gallon for gas, right?
The oil companies don't set gas prices, that's the commodities market. As for doing a great job in the last 5 years they've boosted production from 85m barrels / day to 90m against a world where easy to get to reserves are collapsing. Yes they are doing a great job. They are making far too much money for that great job, but gas would be $75/gallon if it wasn't for the technologies they introduced in the last generation.
If that were true, then care to explain how we came to define and follow Moores law for many a decade?
What does the transistor density have to do with signal error rates?
Sorry, but an unlimited data plan is not some sort of physical impossibility.
Actually there are limits. That was his point. As for unlimited internet in other markets, generally those are capped to. But over the air is vastly more complex than wired so you bump against the one limit and not the other. People who backup their entire Blu-Ray collection notice the home internet limits.
If all phones would default to using wifi for everything whenever it is available then it would take a huge chunk of the load off of the cell towers.
The iPhone does that. I don't know that the carriers want to force this though. And no this is not artificial. The real problem is congress not cannibalizing the HDTV bandwidth for over the air internet. Who uses rabbit ears anymore?
Great post, and witty!