Lets say X releases their software (x) under the GPL.
1) Company Y has software that tightly combines with (x). That code has to be released under the GPL. 2) Company Y writes software (y) and releases it under the GPL. They also write some more software (z) which combines with (y). They release y+z. In this case they can do anything they want.
Well excuse me for thinking about what's best for the people who own the spectrum.
What's best for the people who own the spectrum is that things run smoothly for the carriers so they can pay treasury huge license fees for lots of spectrum.
I haven't seen that at all. People who are getting sued are generally getting sued for violating the GPL and on purpose. That's not lawyer parasites but the kinds of lawsuits that happen with commercial software all the time.
Back then, we had people inventing things like the web
Which took decades, GML started around in 1960. ISIL was in the 1980s. That's not a fair comparison you have no idea what technologies being invented today are important for the computing world of 2030. How would you know?
I can tell you as someone who was around when the web starting being used in the early 1990s I didn't think of it as all that big a deal. I actually thought Gopher with built in indexing was going to be better than the HTML with graphics.
But how groundbreaking is that diversity?
Pretty huge. Most compilers even 10 years ago were heavily optimized for one type of code (language) to one type of CPU in one type of configuration. Even slight changes to hardware required massive shifts in the compiler, often essentially a rewrite. Today we have multi-stage compilers doing very complex compiles with a few man years at most going into getting any particular piece of hardware to work.
In terms of functional compilers,.NET being the best example there was nothing like that a decade ago in mainstream use. In terms of the areas that compilers are moving (again.NET being a leader) with things like Hindley–Milner inference or tail-recursive reduction becoming standard those are huge improvements.
I could go into details but this was just one example of improvement. Another area is the whole high speed handling of video, which is similar to the work that happened on sound in the 1990s.
We are losing each other I think, or at least I'm losing you. I read up the thread I'm not seeing anything about data. So can you clarify what at this point I'm responding to.
From my perspective the GP was just giving an argument for LLVM (it is BSD licensed and thus a better fit for the BSD culture) but phrased this more generally, you questioned him and I said he was over generalizing you then agreed. Which is all good, except there is something about data here...?
However, granting all users patents that may otherwise accidentally get violated is nothing but an act of bad faith,
What it is an act of is an attempt to use copyright law to address a problem in black letter patent law. A patent creates a way to violate the spirit of GPLv2 without violating the letter. The purpose of the GPLv3 is to close these kinds of loopholes.
and the termination clauses, which would allow any author to revoke their license in the event of a violation, is what makes it a double edged sword for the user.
Most commercial licenses have termination clauses. Further the GPLv2 likely had an implicit termination clause. If you object to termination clauses you object to US copyright law in general. This isn't something the FSF has any control over.
I think he probably should have said "Building a complete Unix like OS". Unix has always had system compilers and an assumption that even end users would need to compile code.
As far as I understand it this is the approach that LLVM recommends. Using the LLVM GCC front end for your code while you slowly migrate. If you want real GCC you can use the MacPorts version.
They didn't declare a jihad. They indicated that TIVO had exposed a problem with the license and sought to fix it. Particularly since at the time what TIVO was doing was likely to become more mainstream with things like the Palladium initiative.
Given that track record, which company in its right mind, even if they endorsed the liberation of software, would want to get into bed w/ the FSF?
Companies that want to help a competitor get sued. Or companies that provide something ancillary to free software. 3rd tier players looking to establish niches. But in general few. Very few companies have ever gotten in bed with the FSF through its almost 30 year history. The FSF is rather anti-corporate.
On the other hand lots of companies have successfully used the GPL. It was seen as too strict for many years and then it wasn't.
C which also has a ton of libraries, but was hamstrung by GNU
At the point where Java took off GCC was not a very good compiler and was not a major player. The Linux kernel guys had had to fork it to try and get anything less than terrible performance. Intel's, Watcom, Microsoft, and a 4th I'm forgetting were the big players. In the early 2000s the technical disadvantages of GCC were still rather well known. We are in a rather unusual window for GCC where it seen as not just an open compiler but one of commercial quality.
That doesn't change your point because of course commercial licenses are encumbering as well but factually GCC's GPL status had little if any impact on the mainstream client/server computing of the mid 1990s.
company X writes code for Y that Y sells it as part of a suite to Z. If Y wants to GPL his stuff then X's community can use it. If Y wants to keep his stuff then he had to buy a commercial license from X. This way X got money from people with money, got contributions from people who were willing to contribute and could sue people who did neither.
But company X writes code for Y that Y sells as part of a SaaS suite to Z. doesn't constitute distribution under the GPL. So you either need a stricter license if you want to stop Y's behavior or you might as well use a less strict license if you don't.
Many companies for instance strip all comments from GPL'd source before releasing for legal reason
That's specifically illegal under the GPL. Source code is defined as what is used internally. If a company uses a commented version internally they can be forced to hand that over by the people whose code they are comingling with.
And in this case, hindsight tells us that there are plenty of non-GPL free packages that you use every day that haven't succumbed to either of your fears
Yes but compilers are one area where this sort of thing has happened before and appears to be happening now. Where this fragmentation hasn't happened with BSD projects is when only one entity is heavily invested and cooperation isn't happening or his happening within the one entity.
I'd rather see that again that these modern days where ideas are scarce
When are these modern days with scarce ideas? I'd say we are having a pretty wonderful world of software diversity and experimentation. Just look at the incredible number of web frameworks, for designing interactive websites. Compilers in their hayday never had diversity like this.
I agree with you and good point regarding the X analogy. GCC has benefited time and time again from corporations contributing code that benefit other corporations in uses they never considered. LLVM has some substantial technical advantages over GCC. So we might end up with a situation like we had for a long time where GCC is standard based, feature rich but technically inferior to the commercial compilers many of which are LLVM based.
What about your TiVo? Would make a great copyright infringement device with the right code.
As an aside the TiVo as it ships now includes everything you need for copyright infringement. The DMCA prevents people from selling an "all-in-one" solution for the software side, TIVO itself is not doing anything secret here.
No the BSD does not give the user in practice any freedom. The developer is not, in the Unix non commercial world, the last step between the end user and the software. Those people intermediate in the chain get those rights and then choose based on their own needs whether to pass those rights on or not.
We don't have to guess which model works best, at this point we have historical data. Your model failed with respect to X. MIT created and maintained an X that they released via. the MIT license. All the UNIX vendors then took this MIT code and intermixed it with their custom code creating value add X's that were specific to their platform, and closed source. The effect was that the X that existed in the public domain was worthless for end users, and the X's that were worthwhile were closed. X itself couldn't progress because it fragmented so all the interesting stuff existed in other layers. Years later when there was a desire for a workable open X, the XFree86 project had to start, essentially from scratch and this took years. We still haven't gotten all the features that existed in those proprietary Xs 2 decades ago.
That is the classic example of why BSD style licensing doesn't work. The primary maintainer is not unchanging.
Conversely the GPL has a long history of successful multi corporate contributions over time. The historical data simply refutes your theory of what should work.
They probably want to be able to run simulations that are actually dangerous. Or be able to assume crowd behaviors.
They might want to simulate a terrorist attack or a plague. They might want to use various traffic mitigation strategies with the same traffic jam over and over and over.
JB: What carrier doesn't offer net neutrality at this point. JWR: Then it should be really easy for them to agree to it, right?
Well yeah. It would also be easy for them to agree to having their staff be oxygen breathing. Generally though if you make a demand the demand is for something that the other party either isn't doing or doesn't intend to continue doing.
Red herring. Contracts can take care of making sure the subsidy is paid back - just like they make sure you pay even if you simply cancel and stop using the phone altogether. The point is that some carriers won't unlock phones that are paid off because it suits them to make your phone useless if you want to go to another carrier even though you've fullfilled your part of the bargain.
Collateralized debt and uncollateralized debt have different associated interest rates. That's why a person pays less on a car loan or a mortgage than on credit card debt. By keeping the phone "on their network" the carrier makes sure they capture some (if not all) of the economic advantage of the subsidy regardless of whether you pay. But you are also missing the point about them being "paid back". They still get economic value out of the phone for the life of the phone. Quite often because these phones are frequently given to charities which sell them to the post pay market these phones have economic benefits for 2-3 generations of customers, not just the initial contract.
Lets take for example most Smartphones. Verizon's contract is $350 debt on month 0. With an iPhone 4S they are typically into that phone about $420 (more on the phones with more ram). No business would spend $420 to get a semi-enforceable $350 debt. They are doing it because they want to capture additional economic value. They aren't selling you a phone on some sort of payment plan, they are paying to have a phone on the their network and allowing you complete use/control of the phone in exchange for helping them to offset the cost of introducing that phone to their network.
The same thing applies even moreso with dumb phones where the carrier is getting about $120 in debt for often over $200 in expense.
The point is that some carriers won't unlock phones that are paid off because it suits them to make your phone useless if you want to go to another carrier even though you've fullfilled your part of the bargain
Yes. But you are thinking about the phone from a customer perspective. Regulation is done from an industry perspective. The FCC has no reason to subtly undermine the phone subsidy system. If they disagree with it, they could just ban it. If they agree with it, then to make it work carriers have to continue to get economic benefit even after "you've fulfilled your part of the bargain".
The inexpensive locked phones created for the postpay market, helps to offset costs for the prepay market. This isn't about you it is about the person you sell or give you phone to after you are finished.
It's nearly impossible to roam on other carriers services,
The problem with roaming is that carriers charge each other high roaming fees. So for carriers to offer roaming in any fair / efficient way they have to pass through fees, fees high enough that they sting. Customers hate being charged roaming fees and think their carrier is ripping them off, even if they are just passing through costs.
Why should carriers provide a service from which they often lose money that antagonizes their customers?
What carrier doesn't offer net neutrality at this point. Sometimes their customer facing services which essentially buys services from the wholesale has some forms of internet restrictions to boost revenue and that's an example of troublesome fees and surcharges that are part of American retail pricing. That's not really specific to cell phones. Their wholesale divisions certainly don't care.
As for not carrier locking phones. Phones in the USA are mostly paid for through large subsidies. Why would the government be opposed to carrier locking? Heck why should individuals be opposed to carrier locking?
The carriers would be thrilled. That's what the carriers want. They want to free up parts of the spectrum that provide little benefit (like High Def Television). They are perfectly willing to pay
Wow any idea why Australia & New Zealand have such low caps? I looked at the data and broadband penetration is around where it was in the USA in 2002 or so, so you are about a decade behind in terms of utilization. High prices might account for that.
With caps that low it seems like the telcos are using older technologies like internet over bundled copper (T1 clusters) for the backbones themselves. I get that Australia has terrible population density but I would think you could stil wire up a regional internet all Melborne, Sydney, Brisbane, Canberra which is super fast. In terms of global content the the bandwidth hog is porn and I'm sure there are Australian porn companies that would love to be told they can operate domestically unmolested. For things like YouTube Google could cache a sync a local version...
This seems like a mostly fixable problem to me. Except for Perth and Darwin there is no good reason you should have to pay that much.
Slashdot Mirror
I'm not quite sure what you are saying here.
Lets say X releases their software (x) under the GPL.
1) Company Y has software that tightly combines with (x). That code has to be released under the GPL.
2) Company Y writes software (y) and releases it under the GPL. They also write some more software (z) which combines with (y). They release y+z. In this case they can do anything they want.
Well excuse me for thinking about what's best for the people who own the spectrum.
What's best for the people who own the spectrum is that things run smoothly for the carriers so they can pay treasury huge license fees for lots of spectrum.
I haven't seen that at all. People who are getting sued are generally getting sued for violating the GPL and on purpose. That's not lawyer parasites but the kinds of lawsuits that happen with commercial software all the time.
Back then, we had people inventing things like the web
Which took decades, GML started around in 1960. ISIL was in the 1980s. That's not a fair comparison you have no idea what technologies being invented today are important for the computing world of 2030. How would you know?
I can tell you as someone who was around when the web starting being used in the early 1990s I didn't think of it as all that big a deal. I actually thought Gopher with built in indexing was going to be better than the HTML with graphics.
But how groundbreaking is that diversity?
Pretty huge. Most compilers even 10 years ago were heavily optimized for one type of code (language) to one type of CPU in one type of configuration. Even slight changes to hardware required massive shifts in the compiler, often essentially a rewrite. Today we have multi-stage compilers doing very complex compiles with a few man years at most going into getting any particular piece of hardware to work.
In terms of functional compilers, .NET being the best example there was nothing like that a decade ago in mainstream use. In terms of the areas that compilers are moving (again .NET being a leader) with things like Hindley–Milner inference or tail-recursive reduction becoming standard those are huge improvements.
I could go into details but this was just one example of improvement. Another area is the whole high speed handling of video, which is similar to the work that happened on sound in the 1990s.
We are losing each other I think, or at least I'm losing you. I read up the thread I'm not seeing anything about data. So can you clarify what at this point I'm responding to.
From my perspective the GP was just giving an argument for LLVM (it is BSD licensed and thus a better fit for the BSD culture) but phrased this more generally, you questioned him and I said he was over generalizing you then agreed. Which is all good, except there is something about data here...?
However, granting all users patents that may otherwise accidentally get violated is nothing but an act of bad faith,
What it is an act of is an attempt to use copyright law to address a problem in black letter patent law. A patent creates a way to violate the spirit of GPLv2 without violating the letter. The purpose of the GPLv3 is to close these kinds of loopholes.
and the termination clauses, which would allow any author to revoke their license in the event of a violation, is what makes it a double edged sword for the user.
Most commercial licenses have termination clauses. Further the GPLv2 likely had an implicit termination clause. If you object to termination clauses you object to US copyright law in general. This isn't something the FSF has any control over.
I think he probably should have said "Building a complete Unix like OS". Unix has always had system compilers and an assumption that even end users would need to compile code.
As far as I understand it this is the approach that LLVM recommends. Using the LLVM GCC front end for your code while you slowly migrate. If you want real GCC you can use the MacPorts version.
and declared a jihad on the company
They didn't declare a jihad. They indicated that TIVO had exposed a problem with the license and sought to fix it. Particularly since at the time what TIVO was doing was likely to become more mainstream with things like the Palladium initiative.
Given that track record, which company in its right mind, even if they endorsed the liberation of software, would want to get into bed w/ the FSF?
Companies that want to help a competitor get sued. Or companies that provide something ancillary to free software. 3rd tier players looking to establish niches. But in general few. Very few companies have ever gotten in bed with the FSF through its almost 30 year history. The FSF is rather anti-corporate.
On the other hand lots of companies have successfully used the GPL. It was seen as too strict for many years and then it wasn't.
C which also has a ton of libraries, but was hamstrung by GNU
At the point where Java took off GCC was not a very good compiler and was not a major player. The Linux kernel guys had had to fork it to try and get anything less than terrible performance. Intel's, Watcom, Microsoft, and a 4th I'm forgetting were the big players. In the early 2000s the technical disadvantages of GCC were still rather well known. We are in a rather unusual window for GCC where it seen as not just an open compiler but one of commercial quality.
That doesn't change your point because of course commercial licenses are encumbering as well but factually GCC's GPL status had little if any impact on the mainstream client/server computing of the mid 1990s.
What lawyers have sued over GPL beyond a few special cases?
It is not really an RMS issue.
The model that worked well for GPLv2 was:
company X writes code for Y that Y sells it as part of a suite to Z.
If Y wants to GPL his stuff then X's community can use it.
If Y wants to keep his stuff then he had to buy a commercial license from X.
This way X got money from people with money, got contributions from people who were willing to contribute and could sue people who did neither.
But
company X writes code for Y that Y sells as part of a SaaS suite to Z.
doesn't constitute distribution under the GPL. So you either need a stricter license if you want to stop Y's behavior or you might as well use a less strict license if you don't.
Many companies for instance strip all comments from GPL'd source before releasing for legal reason
That's specifically illegal under the GPL. Source code is defined as what is used internally. If a company uses a commented version internally they can be forced to hand that over by the people whose code they are comingling with.
And in this case, hindsight tells us that there are plenty of non-GPL free packages that you use every day that haven't succumbed to either of your fears
Yes but compilers are one area where this sort of thing has happened before and appears to be happening now. Where this fragmentation hasn't happened with BSD projects is when only one entity is heavily invested and cooperation isn't happening or his happening within the one entity.
I'd rather see that again that these modern days where ideas are scarce
When are these modern days with scarce ideas? I'd say we are having a pretty wonderful world of software diversity and experimentation. Just look at the incredible number of web frameworks, for designing interactive websites. Compilers in their hayday never had diversity like this.
I agree with you and good point regarding the X analogy. GCC has benefited time and time again from corporations contributing code that benefit other corporations in uses they never considered. LLVM has some substantial technical advantages over GCC. So we might end up with a situation like we had for a long time where GCC is standard based, feature rich but technically inferior to the commercial compilers many of which are LLVM based.
What about your TiVo? Would make a great copyright infringement device with the right code.
As an aside the TiVo as it ships now includes everything you need for copyright infringement. The DMCA prevents people from selling an "all-in-one" solution for the software side, TIVO itself is not doing anything secret here.
No the BSD does not give the user in practice any freedom. The developer is not, in the Unix non commercial world, the last step between the end user and the software. Those people intermediate in the chain get those rights and then choose based on their own needs whether to pass those rights on or not.
We don't have to guess which model works best, at this point we have historical data. Your model failed with respect to X. MIT created and maintained an X that they released via. the MIT license. All the UNIX vendors then took this MIT code and intermixed it with their custom code creating value add X's that were specific to their platform, and closed source. The effect was that the X that existed in the public domain was worthless for end users, and the X's that were worthwhile were closed. X itself couldn't progress because it fragmented so all the interesting stuff existed in other layers. Years later when there was a desire for a workable open X, the XFree86 project had to start, essentially from scratch and this took years. We still haven't gotten all the features that existed in those proprietary Xs 2 decades ago.
That is the classic example of why BSD style licensing doesn't work. The primary maintainer is not unchanging.
Conversely the GPL has a long history of successful multi corporate contributions over time. The historical data simply refutes your theory of what should work.
They probably want to be able to run simulations that are actually dangerous. Or be able to assume crowd behaviors.
They might want to simulate a terrorist attack or a plague. They might want to use various traffic mitigation strategies with the same traffic jam over and over and over.
JB: What carrier doesn't offer net neutrality at this point.
JWR: Then it should be really easy for them to agree to it, right?
Well yeah. It would also be easy for them to agree to having their staff be oxygen breathing. Generally though if you make a demand the demand is for something that the other party either isn't doing or doesn't intend to continue doing.
Red herring. Contracts can take care of making sure the subsidy is paid back - just like they make sure you pay even if you simply cancel and stop using the phone altogether. The point is that some carriers won't unlock phones that are paid off because it suits them to make your phone useless if you want to go to another carrier even though you've fullfilled your part of the bargain.
Collateralized debt and uncollateralized debt have different associated interest rates. That's why a person pays less on a car loan or a mortgage than on credit card debt. By keeping the phone "on their network" the carrier makes sure they capture some (if not all) of the economic advantage of the subsidy regardless of whether you pay. But you are also missing the point about them being "paid back". They still get economic value out of the phone for the life of the phone. Quite often because these phones are frequently given to charities which sell them to the post pay market these phones have economic benefits for 2-3 generations of customers, not just the initial contract.
Lets take for example most Smartphones. Verizon's contract is $350 debt on month 0. With an iPhone 4S they are typically into that phone about $420 (more on the phones with more ram). No business would spend $420 to get a semi-enforceable $350 debt. They are doing it because they want to capture additional economic value. They aren't selling you a phone on some sort of payment plan, they are paying to have a phone on the their network and allowing you complete use/control of the phone in exchange for helping them to offset the cost of introducing that phone to their network.
The same thing applies even moreso with dumb phones where the carrier is getting about $120 in debt for often over $200 in expense.
The point is that some carriers won't unlock phones that are paid off because it suits them to make your phone useless if you want to go to another carrier even though you've fullfilled your part of the bargain
Yes. But you are thinking about the phone from a customer perspective. Regulation is done from an industry perspective. The FCC has no reason to subtly undermine the phone subsidy system. If they disagree with it, they could just ban it. If they agree with it, then to make it work carriers have to continue to get economic benefit even after "you've fulfilled your part of the bargain".
The inexpensive locked phones created for the postpay market, helps to offset costs for the prepay market. This isn't about you it is about the person you sell or give you phone to after you are finished.
It's nearly impossible to roam on other carriers services,
The problem with roaming is that carriers charge each other high roaming fees. So for carriers to offer roaming in any fair / efficient way they have to pass through fees, fees high enough that they sting. Customers hate being charged roaming fees and think their carrier is ripping them off, even if they are just passing through costs.
Why should carriers provide a service from which they often lose money that antagonizes their customers?
What carrier doesn't offer net neutrality at this point. Sometimes their customer facing services which essentially buys services from the wholesale has some forms of internet restrictions to boost revenue and that's an example of troublesome fees and surcharges that are part of American retail pricing. That's not really specific to cell phones. Their wholesale divisions certainly don't care.
As for not carrier locking phones. Phones in the USA are mostly paid for through large subsidies. Why would the government be opposed to carrier locking? Heck why should individuals be opposed to carrier locking?
The carriers would be thrilled. That's what the carriers want. They want to free up parts of the spectrum that provide little benefit (like High Def Television). They are perfectly willing to pay
They are perfectly willing to pay very large usage fees to the treasury each year. That's handing value to the public.
Wow any idea why Australia & New Zealand have such low caps? I looked at the data and broadband penetration is around where it was in the USA in 2002 or so, so you are about a decade behind in terms of utilization. High prices might account for that.
With caps that low it seems like the telcos are using older technologies like internet over bundled copper (T1 clusters) for the backbones themselves. I get that Australia has terrible population density but I would think you could stil wire up a regional internet all Melborne, Sydney, Brisbane, Canberra which is super fast. In terms of global content the the bandwidth hog is porn and I'm sure there are Australian porn companies that would love to be told they can operate domestically unmolested. For things like YouTube Google could cache a sync a local version...
This seems like a mostly fixable problem to me. Except for Perth and Darwin there is no good reason you should have to pay that much.