On the other hand if “Minix” becomes synonymous with “that spyware OS” in public perception maybe it will get Tannenbaum to realize the whole “BSD licensing offers the users maximum freedom” schtick is hilariously backwards.
Tannenbaums statement about the news MINIX was being used continued to support the BSD license. He only wished that Intel had told him, cause it would have been cool to know.
He also said “this bit of news reaffirms my view that the Berkeley license provides the maximum amount of freedom to potential users” which is just utterly delusional.
Intel is the user. Intel got maximum freedom. I don’t see what is ambiguous about that.
Unless, are you are recursively applying a broader concept of freedom to all users potentially affected by consequences of the license choice? Because that’s an entirely separate matter altogether. And highly subjective.
It seems to me that I am the user of my CPU and all the software on it, Intel is just the manufacturer. At least legally, morally, and in terms of how GP used the term. Unfortunately there may be an argument that Intel is a user in reality.
You’re a user of your CPU as a complete product, not of MINIX. Intel is the user of MINIX, they are the ones the license applies to. Legally speaking, that much is completely unambiguous.
Intel is the distributor of Minix, they require a license to create copies and distribute it because that’s how copyright works.
I do not need a license for my use of Minix because my use does not create a copy or otherwise implicate copyright law. That does not mean I am not the user, legally or otherwise. Just as I am the user of a book when I read it, but I do not require a license to do so.
Hmmm. Just like I may be a user of a Dell Machine with Windows on it? But in that case, the license applies to me.
MINIX’s license applies equally to source and binary forms, so the “user” is actually quite questionable. Technically, it seems that the license is passed on to me, the user of the CPU, since the copyright notice and such must be distributed along with a binary distribution…
So, I guess I could legally redistribute my changes to the binary blob that Intel puts on the CPU, too, no? Of course, provided my changes are also released under a BSD license. (Difficulty of this aside)
Hmmm. Just like I may be a user of a Dell Machine with Windows on it? But in that case, the license applies to me.
Totally different. Intel is licensing MINIX as a component. Dell is an authorized reseller of Windows, i.e. they are authorized to sell you a Windows license along with their product. While they probably license Windows for their employees, they don’t license Windows as a part of the product. Their terms of resale allow them to modify the original distribution, to include drivers and so on. Also crapware, apparently.
MINIX’s license applies equally to source and binary forms, so the “user” is actually quite questionable.
No, it isn’t. The user of the license is the distributor. The license notice must be attached. That’s all.
Technically, it seems that the license is passed on to me, the user of the CPU, since the copyright notice and such must be distributed along with a binary distribution…
A notice of license use does not constitute a license grant. That is, if I buy a software library A and include it in my product, that does not transitively grant a license for A to all of my users. That would be absurd.
So, I guess I could legally redistribute my changes to the binary blob that Intel puts on the CPU, too, no? Of course, provided my changes are also released under a BSD license.
BSD does not require you release derivatives under a BSD license, only that you attach the original license in the documentation or other materials provided with your derivation. And you could only redistribute changes if Intel’s license for their blob permits redistribution.
He also said “this bit of news reaffirms my view that the Berkeley license provides the maximum amount of freedom to potential users” which is just utterly delusional.
How is it delusional exactly? Seems to be correct, the freedom for intel to use it in the way they see fit is preserved.
BSD is about developer freedom whereas GPL is about end user freedom. For Tannenbaum to claim otherwise is bizarre. He should be arguing for the license he chose on its merits.
What are you talking about? He said BSD provides maximum freedom to potential users. And the next sentence makes it extremely clear he is referring to developers like Intel, not end users.
So as far as I’m aware Tannenbaum never made any such claim. He certainly didn’t in his open letter to Intel.
Intel is the user here, and Intel indeed got maximum freedom. I see nothing wrong with that.
What kind of license would you suggest?
GPL/copyleft? Wouldn’t do anything, Intel would just reluctantly post an uninteresting dump of MINIX source with the tiny modifications they made to make it fit in a small device.
“Don’t use this for evil” licenses? They’re just broken. Incompatible with any existing FOSS licenses. No one will use your code.
The whole point of the tivoization clauses of the GPLv3 is to make it illegal to prevent the end user from modifying this kind of thing and allowing them to remove user-hostile behavior.
BSD forbids the use of the original name without author’s permission, so the license is fine. But no license is going to protect you from FUD and journalist misinterpretation.
Awwwwkward….
Why does every headline mention “MINIX based” now? MINIX is not what’s bad about it >_<
On the other hand if “Minix” becomes synonymous with “that spyware OS” in public perception maybe it will get Tannenbaum to realize the whole “BSD licensing offers the users maximum freedom” schtick is hilariously backwards.
Tannenbaums statement about the news MINIX was being used continued to support the BSD license. He only wished that Intel had told him, cause it would have been cool to know.
He also said “this bit of news reaffirms my view that the Berkeley license provides the maximum amount of freedom to potential users” which is just utterly delusional.
Intel is the user. Intel got maximum freedom. I don’t see what is ambiguous about that.
Unless, are you are recursively applying a broader concept of freedom to all users potentially affected by consequences of the license choice? Because that’s an entirely separate matter altogether. And highly subjective.
It seems to me that I am the user of my CPU and all the software on it, Intel is just the manufacturer. At least legally, morally, and in terms of how GP used the term. Unfortunately there may be an argument that Intel is a user in reality.
You’re a user of your CPU as a complete product, not of MINIX. Intel is the user of MINIX, they are the ones the license applies to. Legally speaking, that much is completely unambiguous.
Intel is the distributor of Minix, they require a license to create copies and distribute it because that’s how copyright works.
I do not need a license for my use of Minix because my use does not create a copy or otherwise implicate copyright law. That does not mean I am not the user, legally or otherwise. Just as I am the user of a book when I read it, but I do not require a license to do so.
Precisely, Intel is the user of the license. That was the intended meaning of user in Tannenbaum’s letter.
Hmmm. Just like I may be a user of a Dell Machine with Windows on it? But in that case, the license applies to me.
MINIX’s license applies equally to source and binary forms, so the “user” is actually quite questionable. Technically, it seems that the license is passed on to me, the user of the CPU, since the copyright notice and such must be distributed along with a binary distribution…
So, I guess I could legally redistribute my changes to the binary blob that Intel puts on the CPU, too, no? Of course, provided my changes are also released under a BSD license. (Difficulty of this aside)
Disclaimer: IANAL.
Totally different. Intel is licensing MINIX as a component. Dell is an authorized reseller of Windows, i.e. they are authorized to sell you a Windows license along with their product. While they probably license Windows for their employees, they don’t license Windows as a part of the product. Their terms of resale allow them to modify the original distribution, to include drivers and so on. Also crapware, apparently.
No, it isn’t. The user of the license is the distributor. The license notice must be attached. That’s all.
A notice of license use does not constitute a license grant. That is, if I buy a software library A and include it in my product, that does not transitively grant a license for A to all of my users. That would be absurd.
BSD does not require you release derivatives under a BSD license, only that you attach the original license in the documentation or other materials provided with your derivation. And you could only redistribute changes if Intel’s license for their blob permits redistribution.
How is it delusional exactly? Seems to be correct, the freedom for intel to use it in the way they see fit is preserved.
BSD is about developer freedom whereas GPL is about end user freedom. For Tannenbaum to claim otherwise is bizarre. He should be arguing for the license he chose on its merits.
What are you talking about? He said BSD provides maximum freedom to potential users. And the next sentence makes it extremely clear he is referring to developers like Intel, not end users.
So as far as I’m aware Tannenbaum never made any such claim. He certainly didn’t in his open letter to Intel.
Intel is the user here, and Intel indeed got maximum freedom. I see nothing wrong with that.
What kind of license would you suggest?
GPL/copyleft? Wouldn’t do anything, Intel would just reluctantly post an uninteresting dump of MINIX source with the tiny modifications they made to make it fit in a small device.
“Don’t use this for evil” licenses? They’re just broken. Incompatible with any existing FOSS licenses. No one will use your code.
The whole point of the tivoization clauses of the GPLv3 is to make it illegal to prevent the end user from modifying this kind of thing and allowing them to remove user-hostile behavior.
BSD forbids the use of the original name without author’s permission, so the license is fine. But no license is going to protect you from FUD and journalist misinterpretation.