I guess these things are relative. I am glad we do not behave in person as we do on slashdot, or I would have to bring a trauma kit to Linux conferences.:-)
One license with a special notification clause isn't that much of a problem. 100 licenses are. Right now, the volunteer hacking on free software knows a few rules of thumb like "distribute the source code" that keep them in compliance with all of our licenses. If they have to get a read from their lawyer on 100 different licenses, each with their special things to do: send a postcard, post to a web site, etc., it's going to become very difficult for that volunteer hacker.
So, I guess you could say I'm fighting creeping complex-ification of licenses.
Geez, guys. I wrote a very polite and non-confrontational letter asking Apple for a few small changes and clarifications. The folks at Apple did not find that unreasonable.
I really wish you'd all stuck with the tone I established. Fortunately, the people who count in this matter seem to understand how to ignore the flames.
You can't stop Slashdot from having a flame war. Too much unutilized testosterone around here:-)
My letter was non-confrontational, and it was taken that way by the people it was addressed to. We had a little laugh together about the inevitability of us all being chopped to giblets on Slashdot.
Nope, read the next paragraph. It says you are still bound by the license in those cases.
The GPL allows you to stop distributing the code if you are bound by court order, etc., it does not say anywhere that you can compel third parties to destroy their copies or stop distributing them.
It's your modifications that are sub-licensed, not the original code. In other words, if you add 10 lines, those 10 lines are sub-licensed.
Bruce
Let me see if I've got this right... (V1.1)
on
OSI APSL Response
·
· Score: 1
If I gave you a license that was "Free until the year 2000", would you accept it as Open Source?
How is a "Free until we take it back" license different, then?
Termination implicitly violates the OSD because a terminated license means "All Rights Reserved", and thus fails all 9 tests of the OSD.
If we're going to accept a license with termination at all, it makes sense to be very careful about the conditions in which termination is allowed. Apple seems to feel there is room for tightening this up and making it unambiguous, I look forward to their next license version.
Thanks
Bruce
I spoke with people at Apple today
on
OSI APSL Response
·
· Score: 1
I spoke with the entire Open Source board, but Eric was offline (as usual, he travels almost continually).
Bruce
Speak for yourself, I'm still writing code.
on
OSI APSL Response
·
· Score: 1
In the past several weeks I've ported the Proxim wireless modem drivers to the 2.2 kernel, I've written a script to interface the Mailman mailing-list processor to virtual domains without using aliases, and I've developed a log rotation, compression, and analysis script to drive Webalizer from a large Apache site. All but the last are already released to the public.
Bruce
I spoke with people at Apple today
on
OSI APSL Response
·
· Score: 4
I spoke with Apple's director of operating systems development, and with the specific person responsible for Open Source development at Apple.
They acknowledge that they have not defined Affected Original Code well, and they should do so. That would help with the termination issue. We also discussed the notification issue, and they seemed receptive about that.
One point I try to make with every company is that these licenses are executed by hackers, not lawyers, not even people who have easy access to lawyers. Those hackers have to be able to read the license, and it should be unambiguous to them. The folks at Apple accept that point.
So, at least Apple isn't telling me that I'm blowing smoke. It wouldn't hurt Eric to help ask for these simple changes.
I spoke with people at Apple today - their director of OS development and also the specific person responsible for Open Source development. They welcome suggestions about the license and certainly did not have any problem with my making one. Let's give them some time and see if they can improve the license.
It would still be more pleasing to our eyes if a big fat link to.org was put on.com:)
What would Perens.org be? My own organization for myself?:-) I just can't convince myself that people's domain suffixes mean much any more.
Rather than withdraw the "Bruce is totally nuts" outburst, please think twice next time. I wrote a really even-toned letter and I wish the slashdot argument that followed it had been as even-toned.
I think they are saying that they can't terminate your right to distribute your modifications. However, I have a real attorney at an IP firm who wants to answer these questions for free, so I will discuss them with him.
I sent it to this guy I know at Apple, Steve Jobs. No reply so far. I'll call this VP of Operating System Development who is quoted in the Wired story.
Dspeed.net seems to be having problems, and thus perens.com can not be reached. Here's the document.
Bruce
The Apple Public Source License - Our Concerns Bruce Perens, Primary Author: The Open Source Definition. Co-Founder: The Open Source Initiative. Wichert Akkerman:Debian Project Leader. Ian Jackson: President, Software in the Public Interest. Author, Debian package installation tool `dpkg'.
We welcome Apple Computer, Inc. as a participant in the Free Software Community. We feel that a few problems in the present version of the Apple Public Source License (the APSL) disqualify it as "Open Source(TM)" or "Free Software". We hope that Apple can address these issues to everyone's satisfaction.
The participation of companies like Apple and IBM should be considered in the same way as the participation of any free software developer. Everyone is welcome to make a contribution. Individually, we each decide whether or not to accept a particular developer's contribution, for reasons that range from technical to legal and licensing concerns. We openly discuss these issues before our community, often quite harshly, as a means of developing consensus and charting our course. One consensus that we've reached is the Open Source Definition, a generally accepted definition of Free Software licensing, written by Bruce Perens and the Debian GNU/Linux developers in 1997.
We note that much of the material that Apple has just released under the APSL originated at The University of California, Berkeley and at Carnegie-Mellon University. That work was sponsored by the U.S. Government, paid for with our taxes, and was already available as Free Software under the BSD license and other well-accepted Open Source licenses. Many of these files do not significantly differ from the pre-Apple versions except that they bear the addition of a new copyright and license. Other files are entirely authored by Apple or bear significant modifications that should indeed be considered Apple's property. Where Apple has not significantly modified individual files from their pre-Apple versions, their original licenses should be preserved without the addition of the APSL.
Section 2.2(c) of the APSL requires that the producer of modifications to APSL-licensed code use a particular URL in the Apple.com domain to notify Apple. While the demise of Apple Computer, Inc. is unlikely in the near future, that sad event would leave us unable to comply with this section of the APSL. This would constitute a restriction on all rights granted by the license, including those rights necessary to qualify under the Open Source Definition. The Free Software community plans a very long lifetime for its software, and we hope that Apple will cooperate by changing this provision so that APSL-licensed software could survive without Apple. We suggest that the simple publication of modifications, such as posting on a personal web site accessible to the global internet and pointed out in any binary distributions, be all that is required. This is consistent with other licenses in our community.
Section 9.1 of the APSL allows Apple to terminate our rights to use any or all APSL-covered code, at its sole discretion, in the event of an unproven claim of infringement, no matter how specious. This is derived from a similar objectionable portion of IBM's Jikes license, which disqualified that license from being referred to as "Open Source". We hope that Apple will consider the investment that members of the Free Software community will put into APSL-licensed code when they write modifications for it. An arbitrary termination could cause us to suddenly lose that investment at some future date, with no chance for appeal. The licenses accepted by our community do not provide the possibility of termination in this manner. If termination due to an infringement claim is to be allowed at all, it should be explicitly limited to the particular source-code lines that are considered to infringe upon an existing patent. This would make it possible for the free software community to "write around the problem" and create a non-infringing version. The authors of the APSL apparently did not consider that patents expire. It should be possible for us to store infringing code for restoral to use upon the expiration of the patent in question. Apple might also consider if it's possible to allow third-parties to defend the disputed code from an infringement claim that would cause us all to lose our rights under the APSL.
We also regret to note that that Eric Raymond, with the best of intentions, jumped a little too fast to embrace the APSL in his enthusiasm to welcome Apple to our community. He placed the Open Source designation on a license that wasn't quite ready for that. We invite Eric and other members of the Free Software community to join us in requesting the few simple changes to the APSL that we have outlined in this letter.
Contact: Bruce Perens <bruce@perens.com> 510-526-1165 (USA) Links to Relevant Information
Apple added its license and copyright without removing the other licenses and copyrights that had already been applied to the files (in most cases the BSD license). So they have not plagiarized, but you still have to comply with all of the licenses on the file just because Apple stuck a notice on it.
I left after Eric overruled the entire board on another issue. From my discussions with them, it does not appear my vote would have mattered in this issue, and had I remained on the board I would not have been free to comment on the APSL as I have.
Slashdot Mirror
So, it's time for the KDE-haters to stop.
Like the GPL, Qt is free for use with free software only. Proprietary software needs a different license.
Gnome has its libraries LGPL-ed, and is thus OK for both free and proprietary software.
I still like Gnome and run it at home, but there is nothing wrong with Qt's license now.
Thanks
Bruce
I guess these things are relative. I am glad we do not behave in person as we do on slashdot, or I would have to bring a trauma kit to Linux conferences. :-)
So, I guess you could say I'm fighting creeping complex-ification of licenses.
Thanks
Bruce
I really wish you'd all stuck with the tone I established. Fortunately, the people who count in this matter seem to understand how to ignore the flames.
Thanks
Bruce
Thanks
Bruce
My letter was non-confrontational, and it was taken that way by the people it was addressed to. We had a little laugh together about the inevitability of us all being chopped to giblets on Slashdot.
Thanks
Bruce
The GPL allows you to stop distributing the code if you are bound by court order, etc., it does not say anywhere that you can compel third parties to destroy their copies or stop distributing them.
Thanks
Bruce
Bruce
How is a "Free until we take it back" license different, then?
Termination implicitly violates the OSD because a terminated license means "All Rights Reserved", and thus fails all 9 tests of the OSD.
If we're going to accept a license with termination at all, it makes sense to be very careful about the conditions in which termination is allowed. Apple seems to feel there is room for tightening this up and making it unambiguous, I look forward to their next license version.
Thanks
Bruce
Bruce
Bruce
They acknowledge that they have not defined Affected Original Code well, and they should do so. That would help with the termination issue. We also discussed the notification issue, and they seemed receptive about that.
One point I try to make with every company is that these licenses are executed by hackers, not lawyers, not even people who have easy access to lawyers. Those hackers have to be able to read the license, and it should be unambiguous to them. The folks at Apple accept that point.
So, at least Apple isn't telling me that I'm blowing smoke. It wouldn't hurt Eric to help ask for these simple changes.
Thanks
Bruce
Thanks
Bruce
Bruce
Bruce
What would Perens.org be? My own organization for myself? :-) I just can't convince myself that people's domain suffixes mean much any more.
Rather than withdraw the "Bruce is totally nuts" outburst, please think twice next time. I wrote a really even-toned letter and I wish the slashdot argument that followed it had been as even-toned.
Thanks
Bruce
Thanks
Bruce
Bruce
Bruce
What a headache, though. I was on musical hold for hours.
Bruce
Why did I get perens.com? So nobody else would get it and use it to my detriment. That's all.
I worked there 12 years and left to form another company. 12 years is a long time to be at one company.
Bruce
The Apple Public Source License - Our Concerns Bruce Perens , Primary Author: The Open Source Definition. Co-Founder: The Open Source Initiative.
Wichert Akkerman: Debian Project Leader.
Ian Jackson: President, Software in the Public Interest. Author, Debian package installation tool `dpkg'.
We welcome Apple Computer, Inc. as a participant in the Free Software Community. We feel that a few problems in the present version of the Apple Public Source License (the APSL) disqualify it as "Open Source(TM)" or "Free Software". We hope that Apple can address these issues to everyone's satisfaction.
The participation of companies like Apple and IBM should be considered in the same way as the participation of any free software developer. Everyone is welcome to make a contribution. Individually, we each decide whether or not to accept a particular developer's contribution, for reasons that range from technical to legal and licensing concerns. We openly discuss these issues before our community, often quite harshly, as a means of developing consensus and charting our course. One consensus that we've reached is the Open Source Definition, a generally accepted definition of Free Software licensing, written by Bruce Perens and the Debian GNU/Linux developers in 1997.
We note that much of the material that Apple has just released under the APSL originated at The University of California, Berkeley and at Carnegie-Mellon University. That work was sponsored by the U.S. Government, paid for with our taxes, and was already available as Free Software under the BSD license and other well-accepted Open Source licenses. Many of these files do not significantly differ from the pre-Apple versions except that they bear the addition of a new copyright and license. Other files are entirely authored by Apple or bear significant modifications that should indeed be considered Apple's property. Where Apple has not significantly modified individual files from their pre-Apple versions, their original licenses should be preserved without the addition of the APSL.
Section 2.2(c) of the APSL requires that the producer of modifications to APSL-licensed code use a particular URL in the Apple.com domain to notify Apple. While the demise of Apple Computer, Inc. is unlikely in the near future, that sad event would leave us unable to comply with this section of the APSL. This would constitute a restriction on all rights granted by the license, including those rights necessary to qualify under the Open Source Definition. The Free Software community plans a very long lifetime for its software, and we hope that Apple will cooperate by changing this provision so that APSL-licensed software could survive without Apple. We suggest that the simple publication of modifications, such as posting on a personal web site accessible to the global internet and pointed out in any binary distributions, be all that is required. This is consistent with other licenses in our community.
Section 9.1 of the APSL allows Apple to terminate our rights to use any or all APSL-covered code, at its sole discretion, in the event of an unproven claim of infringement, no matter how specious. This is derived from a similar objectionable portion of IBM's Jikes license, which disqualified that license from being referred to as "Open Source". We hope that Apple will consider the investment that members of the Free Software community will put into APSL-licensed code when they write modifications for it. An arbitrary termination could cause us to suddenly lose that investment at some future date, with no chance for appeal. The licenses accepted by our community do not provide the possibility of termination in this manner. If termination due to an infringement claim is to be allowed at all, it should be explicitly limited to the particular source-code lines that are considered to infringe upon an existing patent. This would make it possible for the free software community to "write around the problem" and create a non-infringing version. The authors of the APSL apparently did not consider that patents expire. It should be possible for us to store infringing code for restoral to use upon the expiration of the patent in question. Apple might also consider if it's possible to allow third-parties to defend the disputed code from an infringement claim that would cause us all to lose our rights under the APSL.
We also regret to note that that Eric Raymond, with the best of intentions, jumped a little too fast to embrace the APSL in his enthusiasm to welcome Apple to our community. He placed the Open Source designation on a license that wasn't quite ready for that. We invite Eric and other members of the Free Software community to join us in requesting the few simple changes to the APSL that we have outlined in this letter.
Contact: Bruce Perens <bruce@perens.com> 510-526-1165 (USA) Links to Relevant Information
Thanks
Bruce
Thanks
Bruce