Woz's influence is rather limited after 1980. Apple at the time Woz was involved was one of many really interesting hobbyist computer companies. It was not a major center of innovation. And Steve Jobs did not screw him over. He got hurt, wanted a demotion, and then quit. The only reason people even know who Steve Wozniak is is because of Steve Jobs.
I don't subscribe to the idea that you should be allowed to be anti-competitive as long as you are not a monopoly, because if something was wrong for one company to do then surely it is just as wrong for every company to do.
That's not what a violation of anti-trust law means. What the law prevents is using a monopoly in one area to extend to a monopoly in another. There is nothing illegal about being anti-competitive what they are concerned about is maintaining a diverse marketplace where a natural monopoly hasn't formed.
I know a lot of people who think office for Mac is better than office for Windows. I don't agree, I love things like Access, but the fact someone can plausibly make the claim....
NPD, you have to pay so I can't deep link. They first measured it at 91% in June 2009. This is when Apple's market share of all laptops was only 8.7%, it is over 10% today.
In terms of anti-trust they already make it easy to install Windows and Linux. I guess the court could demand they offer Windows versions. But really I think the argument would be their operating system.
I guess I could see a court forcing them to sell their operating system separately if there was a compelling interest in high end PC laptops. But the problem is there is no reason to believe that this market is being stifled.
I agree that it is too early. This one is going to hurt me personally. Even my wife, is unhappy with this change she loves the reliability of wired. We are both going to get adapters.
That being said let me explicate the why someone could be in favor of this. To start I'm not sure where you are getting a 1/16th of an inch the difference between the air and the regular it is about 1/4 inch. But it is not the size that matters it is the drop in weight from 4.5 lbs to 2.9 lbs (for 13 inch). Weight really matters especially to women.
Apple has a single multifunction port that everything could be attached to, the Thunderbolt port. Thunderbolt + USB is a good direction for the best aesthetic experience. A single port you plug into with monitor, keyboard, mouse (or likely external trackpad), tablet, high speed local storage, internet. Longer term, even potentially extra CPU and memory. Your computer goes from a travel model, to a desktop model instantly.
Apple hasn't been able to move in this direction because while they've brought out the Thunderbolt they haven't really been pushing the hardware standard. Their monitor is not ubiquitous in offices for example. If Apple has decided they are going to move the platform in this direction, they are going to force all the other changes. The net effect is likely to be strongly positive. This is typical of Apple's approach, decide on a direction and move the platform.
I understand how you are reasoning through your essay. But no, that is not what happened at all. This is like one of those fiction pieces where you have people from a different time period responding to the events, like Liberty Kids which has kids with a totally modern mindset during the revolutionary war.
People back in the early to mid 1990s didn't think about compiler licenses much. For several decades compilers all had licenses which, unless you were writing a compiler or an OS, let you do what you wanted. Licensing question just didn't come up for business programmers. I certainly was aware of GPL issues, but I was an exception.
GCC was LGPL so again code generated from GCC was fine. Moreover if you intended to distribute your application you would have used a commercial compiler not GCC. GCC was a great little compiler to throw on Unix boxes where you didn't want to pay the $2k for CC, but still wanted to compile some stuff. If there was actual money involved you used CC or one of the others.
In terms of licensing, it was when companies starting wanting to redistribute their entire solution, including the Linux kernel or MySQL or applications written against KDE that GPL code problems started to emerge and be talked about. That wouldn't happen for years.
Additionally the most popular compiler for Java, the one I used, was Microsoft's Java, later J++. This was part of visual studio. So even with Java most people didn't want free tools. Sun still charged for most of their their tools as well. People were used to paying for compilers and development tools. There were free languages which were popular on Unix and with Unix system administrators. But Unix developers still mostly used paid languages. So when you talk about user pools you have people who generally consider application development languages to be paid while "scripting" is free, which is a different mindset from today. These people knew about GCC, but GCC was unquestionably inferior from a quality standpoint to commercial C compilers.
If you want to get into the correct mindset you need to think of Linux as a cheap bad version of Solaris. It ran on x86 hardware which was ubiquitous and dollar per processing unit incredibly cheap but quality wise well below everything else out there.
In the early 1990s Unix had fragmented terribly. C programs developed in house quite often had hardware dependencies so even if you had the source, porting them from Unix to Unix was complex, time consuming and expensive. This resulted in vendor lockin for servers. The main advantage of GCC that is was that it allowed open source applications to target one particular compiler, combined with a portable operating system this offered the possibility of portability across hardware for C. The big advantage with Java was binary compatibility across different hardware vendors. There were 3 main solutions being considered at the time:
a) Use whatever server hardware you wanted along with Java which promised a common virtual machine. Write once, run anywhere. b) Use x86 hardware along with Microsoft's WindowsNT, which was starting to become viable on the server level. And with the rapid improvements in x86 hardware would become a more viable target every year. c) Use x86 hardware with a free operating system and LAMP. Assume you will be able to cross hardware boundaries along with Linux if the need arises.
Java was part of a hardware strategy not a licensing issue. No one cared about licensing. They didn't care that much about free tools. C failed because it didn't allow for the migration from Unix to Windows and Linux/BSD, C porting for business application was too difficult. Sorry to disagree so strongly with your theory.
Yes you cannot remove any lines from the version you worked on, the preferred form for modification that was used to create the binary you are distributing. I don't know what you mean by "original" there 3 different originals in the source code variant I was responding to.
Apple is not Dell. They aren't into "build you own your way". They design quality balanced systems. They drive their entire platform which reduces complexity for software vendors and users. If you want something outside the norm, they have wonderful solution you just pay a premium.
They are not into maximum choice, they are into good choices.
Apple controls 90% of the over $1000 laptop market. It is core to their brand identity. It is a huge source of products. Further their entire development platform is dependent on OSX. Yes they make a lot more money on phones. They actually make very little on tablets and don't really have a good strategy for higher tablets.
I don't usually agree with you. This one I do. Losing physical ethernet is a series downgrade. Being able to plug in physically is a huge advantage when things go wrong or when you want to deal with network hardware. I'm sure there will be some sort of attachment I have to buy and yeah this is going to cost me.
The GPL is clearly making a distinction between high level "source code" and other forms of the code, so someone can't offer up, say, an intermediate assembly language output.... As I correctly pointed out, if the "conveyed" binary was compiled from a source with no comments in it, then source with no comments is all one has to provide
No that's not true, that is not what the GPL is doing. Source code is not defined as what you compile from but what you modify from. Source code the way you are using it, is explicitly excluded from the license. It is eliminating all intermediate codes including stripped source.
This in fact came up in European courts where Microsoft provided some complex XML quasi-binary code and the courts tossed it as meeting the definition of source, using the GPL definition that source was what Microsoft works in.
I prefer the form of the work which has a binding contract paying me $1M for viewing the source. I demand you provide source code.
That would be a license fee and other licenses include license fees all the time.
You even quoted the correct line, "The 'source code' for a work means the preferred form of the work for making modifications to it" The version used by developers for making modifications is the one with the comments.
As for the GPL what prevents someone from stripping comments is that the recipient can request source.
Everyone including Theo de Raadt would agree that Theo is much harsher that Richard. No one for example has ever believed that Richard Stallman was likely to get violent.
He's always making rather critical quotes like, "Linux people do what they do because they hate Microsoft. We do what we do because we love Unix." For that matter his comment to Richard, "...you are being the usual slimy hypocritical asshole... You may have had value ten years ago, but people will see that you don't anymore."
You might want to check out:
1) NetBSD vs. OpenBSD wars 2) Theo's attacks on FreeBSD and NetBSD's privacy policies 3) Theo's attacks on FreeBSD having GPL code in the kernel. 4) Theo's attacks on NetBSD maintaining legacy hardware too long 5) Theo's attacks on HP being cheap 6) Theo's attacks on Linus programming in the kernel
Linus has had those opinions for about 15 years maybe longer. That being said in 2005 Linus released GIT, and could have picked any license. He picked the GPL not a BSD license. He's been a strong advocate for the GPLv2 saying it was the best choice he ever made in terms of forcing contributions back. The Linux kernel has corporate contributors by the boatload.
I think it is fair to say that Linus is very much on the GPL side of the debate while objecting to Stalman's tactics.
Apple has just on this version of iOS made the MacRuby distribution included in system directories. They are close to making MacRuby their standard for dynamic language application development.
Nor is there an equivalent of the FSF in any of the BSDs
Sure there is, OpenBSD.
One thing I'm curious about - has anybody ever stayed away from BSD on the grounds that they are Unix, and have fears that they may contain proprietary code of USL/SCO/AT&T/UCB/BSDI?
Yes in the early 1990s there was a problem. It was resolved but it happened during the crucial time period around 1992-4 when the world could have gone either way (in most BSD's guy's opinion, I don't agree).
it's amazing that people still spread this kind of fearmongering despite the fact that this scenario has never come true.
Of course it has. The classic case was X. The versions of X people cared about were OS specific versions. The free MIT version existed but was mostly worthless. When someone wanted to create an X, with the XFree86 project they pretty much had to start from scratch. The experience with X was what drove people toward the GPL.
Huh? LLVM came from University if Illinois in 2004. It came from research consideration that had nothing to do with concern about the GPL. Apple picked it up because of concern about the GPL. GCC has been GPL since it started. BSD used to use ICC (generally under SCO) and cross compile. Moving to GCC was moving away from a commercial OS.
And the KDE guys had problems with QT when it was under the QPL. They agreed with Debian legal's analysis that redistribution for anyone other than KDE itself was legally questionable.
Just adding to this... Apple in fact actually does release the code back for almost everyone of their BSD projects. They don't tend to take these things private.
I have to bet my business all the time I'm not going to get sued for far worse long legal documents written by big companies. The nice thing with GPLv3 is that most of the people who hold the license don't actually have that much money.
Slashdot Mirror
Woz's influence is rather limited after 1980. Apple at the time Woz was involved was one of many really interesting hobbyist computer companies. It was not a major center of innovation. And Steve Jobs did not screw him over. He got hurt, wanted a demotion, and then quit. The only reason people even know who Steve Wozniak is is because of Steve Jobs.
I don't subscribe to the idea that you should be allowed to be anti-competitive as long as you are not a monopoly, because if something was wrong for one company to do then surely it is just as wrong for every company to do.
That's not what a violation of anti-trust law means. What the law prevents is using a monopoly in one area to extend to a monopoly in another. There is nothing illegal about being anti-competitive what they are concerned about is maintaining a diverse marketplace where a natural monopoly hasn't formed.
Huh?
http://www.ubuntu.com/download/arm
http://www.debian.org/ports/arm/
http://archlinuxarm.org/
http://www.armedslack.org/
http://maemo.org/
Fedora and Suse are working on an arm version.
All these guys even have a standards committee setup linaro.
I know a lot of people who think office for Mac is better than office for Windows. I don't agree, I love things like Access, but the fact someone can plausibly make the claim....
NPD, you have to pay so I can't deep link. They first measured it at 91% in June 2009. This is when Apple's market share of all laptops was only 8.7%, it is over 10% today.
In terms of anti-trust they already make it easy to install Windows and Linux. I guess the court could demand they offer Windows versions. But really I think the argument would be their operating system.
I guess I could see a court forcing them to sell their operating system separately if there was a compelling interest in high end PC laptops. But the problem is there is no reason to believe that this market is being stifled.
What would be the basis of the complaint?
I agree that it is too early. This one is going to hurt me personally. Even my wife, is unhappy with this change she loves the reliability of wired. We are both going to get adapters.
That being said let me explicate the why someone could be in favor of this. To start I'm not sure where you are getting a 1/16th of an inch the difference between the air and the regular it is about 1/4 inch. But it is not the size that matters it is the drop in weight from 4.5 lbs to 2.9 lbs (for 13 inch). Weight really matters especially to women.
Apple has a single multifunction port that everything could be attached to, the Thunderbolt port. Thunderbolt + USB is a good direction for the best aesthetic experience. A single port you plug into with monitor, keyboard, mouse (or likely external trackpad), tablet, high speed local storage, internet. Longer term, even potentially extra CPU and memory. Your computer goes from a travel model, to a desktop model instantly.
Apple hasn't been able to move in this direction because while they've brought out the Thunderbolt they haven't really been pushing the hardware standard. Their monitor is not ubiquitous in offices for example. If Apple has decided they are going to move the platform in this direction, they are going to force all the other changes. The net effect is likely to be strongly positive. This is typical of Apple's approach, decide on a direction and move the platform.
I understand how you are reasoning through your essay. But no, that is not what happened at all. This is like one of those fiction pieces where you have people from a different time period responding to the events, like Liberty Kids which has kids with a totally modern mindset during the revolutionary war.
People back in the early to mid 1990s didn't think about compiler licenses much. For several decades compilers all had licenses which, unless you were writing a compiler or an OS, let you do what you wanted. Licensing question just didn't come up for business programmers. I certainly was aware of GPL issues, but I was an exception.
GCC was LGPL so again code generated from GCC was fine. Moreover if you intended to distribute your application you would have used a commercial compiler not GCC. GCC was a great little compiler to throw on Unix boxes where you didn't want to pay the $2k for CC, but still wanted to compile some stuff. If there was actual money involved you used CC or one of the others.
In terms of licensing, it was when companies starting wanting to redistribute their entire solution, including the Linux kernel or MySQL or applications written against KDE that GPL code problems started to emerge and be talked about. That wouldn't happen for years.
Additionally the most popular compiler for Java, the one I used, was Microsoft's Java, later J++. This was part of visual studio. So even with Java most people didn't want free tools. Sun still charged for most of their their tools as well. People were used to paying for compilers and development tools. There were free languages which were popular on Unix and with Unix system administrators. But Unix developers still mostly used paid languages. So when you talk about user pools you have people who generally consider application development languages to be paid while "scripting" is free, which is a different mindset from today. These people knew about GCC, but GCC was unquestionably inferior from a quality standpoint to commercial C compilers.
If you want to get into the correct mindset you need to think of Linux as a cheap bad version of Solaris. It ran on x86 hardware which was ubiquitous and dollar per processing unit incredibly cheap but quality wise well below everything else out there.
In the early 1990s Unix had fragmented terribly. C programs developed in house quite often had hardware dependencies so even if you had the source, porting them from Unix to Unix was complex, time consuming and expensive. This resulted in vendor lockin for servers. The main advantage of GCC that is was that it allowed open source applications to target one particular compiler, combined with a portable operating system this offered the possibility of portability across hardware for C.
The big advantage with Java was binary compatibility across different hardware vendors. There were 3 main solutions being considered at the time:
a) Use whatever server hardware you wanted along with Java which promised a common virtual machine. Write once, run anywhere.
b) Use x86 hardware along with Microsoft's WindowsNT, which was starting to become viable on the server level. And with the rapid improvements in x86 hardware would become a more viable target every year.
c) Use x86 hardware with a free operating system and LAMP. Assume you will be able to cross hardware boundaries along with Linux if the need arises.
Java was part of a hardware strategy not a licensing issue. No one cared about licensing. They didn't care that much about free tools. C failed because it didn't allow for the migration from Unix to Windows and Linux/BSD, C porting for business application was too difficult. Sorry to disagree so strongly with your theory.
Yes you cannot remove any lines from the version you worked on, the preferred form for modification that was used to create the binary you are distributing. I don't know what you mean by "original" there 3 different originals in the source code variant I was responding to.
Apple is not Dell. They aren't into "build you own your way". They design quality balanced systems. They drive their entire platform which reduces complexity for software vendors and users. If you want something outside the norm, they have wonderful solution you just pay a premium.
They are not into maximum choice, they are into good choices.
Apple controls 90% of the over $1000 laptop market. It is core to their brand identity. It is a huge source of products. Further their entire development platform is dependent on OSX. Yes they make a lot more money on phones. They actually make very little on tablets and don't really have a good strategy for higher tablets.
So no, this is not going to happen.
If you have a stationary machine they still sell the macpro. This is about the laptops.
I don't usually agree with you. This one I do. Losing physical ethernet is a series downgrade. Being able to plug in physically is a huge advantage when things go wrong or when you want to deal with network hardware. I'm sure there will be some sort of attachment I have to buy and yeah this is going to cost me.
The GPL is clearly making a distinction between high level "source code" and other forms of the code, so someone can't offer up, say, an intermediate assembly language output.... As I correctly pointed out, if the "conveyed" binary was compiled from a source with no comments in it, then source with no comments is all one has to provide
No that's not true, that is not what the GPL is doing. Source code is not defined as what you compile from but what you modify from. Source code the way you are using it, is explicitly excluded from the license. It is eliminating all intermediate codes including stripped source.
This in fact came up in European courts where Microsoft provided some complex XML quasi-binary code and the courts tossed it as meeting the definition of source, using the GPL definition that source was what Microsoft works in.
I prefer the form of the work which has a binding contract paying me $1M for viewing the source. I demand you provide source code.
That would be a license fee and other licenses include license fees all the time.
You even quoted the correct line, "The 'source code' for a work means the preferred form of the work for making modifications to it" The version used by developers for making modifications is the one with the comments.
As for the GPL what prevents someone from stripping comments is that the recipient can request source.
Everyone including Theo de Raadt would agree that Theo is much harsher that Richard. No one for example has ever believed that Richard Stallman was likely to get violent.
He's always making rather critical quotes like, "Linux people do what they do because they hate Microsoft. We do what we do because we love Unix." For that matter his comment to Richard, "...you are being the usual slimy hypocritical asshole... You may have had value ten years ago, but people will see that you don't anymore."
You might want to check out:
1) NetBSD vs. OpenBSD wars
2) Theo's attacks on FreeBSD and NetBSD's privacy policies
3) Theo's attacks on FreeBSD having GPL code in the kernel.
4) Theo's attacks on NetBSD maintaining legacy hardware too long
5) Theo's attacks on HP being cheap
6) Theo's attacks on Linus programming in the kernel
etc...
Linus has had those opinions for about 15 years maybe longer. That being said in 2005 Linus released GIT, and could have picked any license. He picked the GPL not a BSD license. He's been a strong advocate for the GPLv2 saying it was the best choice he ever made in terms of forcing contributions back. The Linux kernel has corporate contributors by the boatload.
I think it is fair to say that Linus is very much on the GPL side of the debate while objecting to Stalman's tactics.
Apple has just on this version of iOS made the MacRuby distribution included in system directories. They are close to making MacRuby their standard for dynamic language application development.
Ruby is GPL.
Nor is there an equivalent of the FSF in any of the BSDs
Sure there is, OpenBSD.
One thing I'm curious about - has anybody ever stayed away from BSD on the grounds that they are Unix, and have fears that they may contain proprietary code of USL/SCO/AT&T/UCB/BSDI?
Yes in the early 1990s there was a problem. It was resolved but it happened during the crucial time period around 1992-4 when the world could have gone either way (in most BSD's guy's opinion, I don't agree).
it's amazing that people still spread this kind of fearmongering despite the fact that this scenario has never come true.
Of course it has. The classic case was X. The versions of X people cared about were OS specific versions. The free MIT version existed but was mostly worthless. When someone wanted to create an X, with the XFree86 project they pretty much had to start from scratch. The experience with X was what drove people toward the GPL.
Oh the reasons for portable compilers and operating systems show up a lot.
1) Most end user programs don't require low level access to hardware.
2) "Best hardware" changes rapidly, much more rapidly than software life-cycles.
Thus the intermediate levels should abstract hardware.
Both GCC and LLVM are very portable as compilers go. Both of them support operating systems that are widely cross platform.
And the KDE guys had problems with QT
Sorry should be: And the BSD guys had problems with QT...
Huh? LLVM came from University if Illinois in 2004. It came from research consideration that had nothing to do with concern about the GPL. Apple picked it up because of concern about the GPL. GCC has been GPL since it started. BSD used to use ICC (generally under SCO) and cross compile. Moving to GCC was moving away from a commercial OS.
And the KDE guys had problems with QT when it was under the QPL. They agreed with Debian legal's analysis that redistribution for anyone other than KDE itself was legally questionable.
Just adding to this... Apple in fact actually does release the code back for almost everyone of their BSD projects. They don't tend to take these things private.
Apple has a long history of contributing back as well as starting open source projects.
http://opensource.apple.com/
http://www.macosforge.org/
1) All the bug fixes from Macports / Darwinports.
2) Darwin itself
3) Bonjour / zeroconfig networking
4) Webkit
5) launchd
6) MacRuby
etc..
I have to bet my business all the time I'm not going to get sued for far worse long legal documents written by big companies. The nice thing with GPLv3 is that most of the people who hold the license don't actually have that much money.