Slashdot Mirror
FSF Statement on Violation of GPL by RTLinux
bkuhn writes "The FSF has issued an official statement on the GPL violation by RTLinux." nothign surprising here, basically they say that RTLinux is violating the GPL by not releasing the source to their Linux kernel mods, but since the FSF isn't the copyright holder, they can't do much about it. Now it's up to RTLinux to decide if they are gonna do the right thing or not.Update: 09/16 00:48 AM GMT by H : Please check out these comments for more information - it's not a source code violation, but a patent issue.
So, Who IS "THE" copyright holder? Linus? A consortium? Would some of the larger distributions be interested in combining together to fight this thing? RedHat/SuSE/Mandrake would do for a start...
I wonder what RTLinux have to say too...
Anyone? Anyone? Bueller? Anyone?
Z.
-- Under/Overrated is meta-moderation, and therefore is Redundant.
...is listed on their site here. Anyone who can translate this into plain English, please do so. IANAL.
Yes. It has always been successfully enforced. It has never come to court though.
-WolfWithoutAClause
"Gravity is only a theory, not a fact!"Specifically, does the original author own the copyright on a large project that other people have made minor contributions to? Also, consider that no one explicitly handed over their copyright.
I think that one of the best ways for GPL developers to make money is by creating software under the GPL and also selling that software under a different license to companies that don't wish to use the GPL. But, I am unclear on whether the original author can do this if other people have made contributions to the project.
I guess an example would be if Torvalds decided to sell Linux to Microsoft under the BSD license. I know, I know - I would shit myself if it happened too. It's just an example.
I've got little sympathy with the guy. Whether his patent is valid or not- the licence he agreed to when he took (took as in stole) the GNU/Linux software and added his patented code to it is pretty clear; and there's descriptions of the GPL all over the net. He could have written his own OS, or distributed patches or closed source modules or device drivers. Or he could still go around and ask every single designer of every single patch to the Linux OS to relicense their code for him to use. Good luck with that.
Now, this version of his software is going to be GPLd. The GNU/Linux designers licensed him their code on that basis. He took it and used it, and he has to pay for what he took. He doesn't have a legal leg to stand on.
Unfair? Probably not. You work out the ratio of the code that he wrote to the code they wrote.
-WolfWithoutAClause
"Gravity is only a theory, not a fact!"My understanding is that on RTLinux there is a ``hard'' real time kernel that runs and that the regular non-realtime Linux kernel and all its processes run as a separate low priority thread of control under this system.
It seems to me that Yodaiken could GPL the mods to the Linux part of the kernel (if he likes) and liscence the real time part of the kernel any way he likes. The patent prevents others from making their own real time kernel and running a non-realtime kernel on top of it, and that idea appears novel (I've not heard of prior art) so he can do with it as he pleases. I don't see how this is any different than a vendor releasing GPL'd device drivers or kernel patches for a patented piece of hardware or software that he would like to support.
Microtest (now XStore) put together a mess of GPL software - a modified Linux kernel 2.0.27, Samba 1.9.x, Apache, the MARS_NWE netware emulator, and GNU C libraries (libc5), among others, stuffed them on a flash chip in a drive-bay-size embedded 486-based computer, and sells it as their "DiscZerver" product line. Nothing wrong with the method, but there's plenty wrong in their implementation.
The web interface, the only given method of configuring the device, refers to the various services installed generically, like "Web server," "SMB server," "NCP server," etc. - there's no mention anywhere, even in the manual, of the actual programs being used. Of course along with this is no accompanying source code or even the offer to provide any, as the GPL requires.
I can't even get any tech support from this company, much less someone to ask about getting the source code for the software and whatever modifications they made, which include a flash-filesystem driver ("yaffs") for the kernel. I did manage to hack out the root password (which they apparently hide from all customers); with that I found a shell prompt (Stand-alone Shell v1.0 - GPL? dunno) which only increased my determination as I could see exactly what programs they managed to steal, strip out identifying info, and use without credit.
I did contact the FSF, and they did confirm the existence of a GPL violation, but were unable to do anything specific as they do not hold copyright on any of these programs (and actually suggested I post to Slashdot to get some answers =] ) Of course xStore itself has not returned my emails or phone call.
So right now I have a nice little piece of hardware, a bunch of GPL software that Microtest 'stole' (for lack of a better word) and no idea what to do next. I'd be happy if I could just get the code so I can fix NMBd to work properly. I've thought about trying to make my own really-small distro to load on, but it's not really worth my time - I could just load the CD images into my other Linux server, connect the CD tower, and get on with life... but I really shouldn't have to do either. Any ideas?
Note:
b) He provides those patches under the GPL (download it and see for yourself)
c) His Patent Licence allows for a blanket Licence, free of charge to all those who incorporate that patents process in GPL'd software.
d) He reserves the right to charge for those incorporating his patent into non-gpl'd software.
So if you hate Software Patents, go ahead and hate Victor (I've talked with him, he's okay with this), however, he is not violating the GPL. He has gone the extra step making his Patent non-violating against the GPL.
Remember that a patent is -not- code. The code is the code. His patches are an implementation of the patent. And his patent licence allows them to be included in GPL'd software without paying him.
If you were to write code that did what the patent described, and did it in a proprietary manner, then you would need to negotiate a licence with Victor, or fight it out with lawyers. You can not like this, as this is the basic software patent bad thing, but he isn't violating the GPL.
Also note the inconsistancies of the FSFs position on Software Patents. Richard has noted that he is for them, if they are used as a pool to force other companies to share thiers, but in this release they say they are completely against them. I'd like to see a public position from Bradly Khun and RMS.
That also said, there are undoubtably a number of places where the Linux Kernel is violating any number of software patents. Get used to seeing these kinds of stories.
Chris DiBona
Co-Editor, Open Sources
Open Source Program Manager, Google, Inc.
IBM Published Proir Art that embodies the same
idea as the RTLinux Patent in releasing in 1967
the CP67 kernal for the Cambridge Monitor System.
CP67, VM and the IBM System/390 Virtual Image
Facility all describe and embody a Real Time
Kernal that runs an entire operating system
(which may be Linux) as a process and that
prevents the client OS (Linux) from disabling
interrupts on the actual CPUs while giving the
appearance of having done so to the client OS.
This is the direct lineal ancestor of the VM/390
and z/VM Operating System and of the System/390
Virtual Image Facility for Linux Kernal that allows
40,000 copies of Linux to run concurrently on a IBM
zSeries mainframe.
Since the Source code for CP67 and all user mods
were published and provided freely to anyone who
had an IBM or compatible Mainframe it could be said
to have been the father of the GPL.
Thanx Doug...