|
Linus Speaks Regarding Claims About Bionic: "It Seems Totally Bogus" - Updated |
|
Tuesday, March 22 2011 @ 02:09 AM EDT
|
Brian Proffitt has the goods, utterly torpedoing the recent claims that Google misused GPL'd header files with Bionic leading to wild predictions of disaster. Proffitt did what Edward Naughton and Florian Mueller could have done prior to making the claims and asked Linus about it directly. Linus told him "it seems totally bogus". The details are fascinating.
Here's the part from Proffitt's article where Linus speaks, but the whole article is worth reading:Naughton raised serious concerns in his report that what Google has done with Bionic (essentially creating a library analogous to glibc for Android) and then licensing Bionic under the Apache license violates the GPL. If that's the case, he argued, then Bionic could be considered a derived work under the GPL and therefore any application that touched Bionic would also be subject to the GPL.
But when I put the question to Linus Torvalds last week, he wasn't nearly as concerned. Torvalds responded to my inquiry just this afternoon and had this to say about the so-called problem:
"It seems totally bogus. We've always made it very clear that the kernel system call interfaces do not in any way result in a derived work as per the GPL, and the kernel details are exported through the kernel headers to all the normal glibc interfaces too.
"The kernel headers contain various definitions for the interfaces to user space, and we even actively try to make sure that the headers can be used by user space (and try to mark which of the headers are expected to be usable in such a way). Exactly because we know user space needs those details in order to interact with the kernel.
"So I haven't looked at exactly what Google does with the kernel headers, but I can't see that they'd want to do anything fundamentally different from glibc in this respect," Torvalds wrote.
Torvalds seems to be at best bemused about the issue and perhaps a tad irritated.
"Of course, we do have our own 'internal' headers too, and we have stuff that is meant to be relevant only for the kernel. But there would be no point for Google to even use those, since they are useless outside of the kernel, so I don't see what the whole brouhaha would be all about. Except if it's somebody politically motivated (or motivated by some need of attention)," he continued. So there you have it. Also, check out
LWN's coverage if you have a subscription, and if you don't I recommend it, but the link has the opening paragraphs, including this significant piece:Way back in the early days of Linux, shortly after Linus Torvalds switched the kernel from his own "non-commercial" license to the GPL, he also added an important clarification to the kernel's license. In the COPYING file at the top of the kernel tree since mid-1993, there has been a clear statement that Torvalds, at least, does not consider user-space programs to be derived from the kernel, and thus are not subject to the kernel's license:
This copyright does *not* cover user programs that use kernel services by normal system calls - this is merely considered normal use of the kernel, and does *not* fall under the heading of "derived work". So, even without talking to Linus, anyone could have known that the claims were over the top. There's an astute comment on that article as well here, that you can read without a sub.
What? You thought if Florian Mueller or some lawyer who has worked for Microsoft pitched an anti-Google/Android claim of doom, it must be so this time? How many times will this happen before journalists stop repeating the FUD and ask the people who know, as Proffitt just did? It's what journalism is about.
Richard Stallman had something to say about header files back in 2003. It's relevant to the FUD that has been making the rounds. Here's what he said: Someone recently made the claim that including a header file always
makes a derivative work.
That's not the FSF's view. Our view is that just using structure
definitions, typedefs, enumeration constants, macros with simple
bodies, etc., is NOT enough to make a derivative work. It would take
a substantial amount of code (coming from inline functions or macros
with substantial bodies) to do that. So. There you have it.
Just because someone makes a speculative claim that Android or Google might be in legal trouble, it doesn't make it true.
And without casting aspersions, I don't care if the person is a lawyer, unless he's a FOSS lawyer. Here's why I say that: from covering the SCO saga, I learned that lawyers who are not FOSS lawyers do not understand the GPL. It works in a way opposite to what they are used to, so it's a supreme puzzlement to them. Remember SCO's "The GPL is Unconstitutional" gambit?
People laughed so hard at them for that, they
dropped it. I'm smiling just remembering it. But real lawyers from a famous law firm came up with that. That didn't make it so. The header files story was just the latest example of a recent flood of "The Google/Android Sky Is Falling" nonsense. Somebody reeeeally wants us to stop using Android, and it's driving them crazy because the market has simply fallen in love with Android, despite their tireless FUD. So, here's my advice to journalists and everyone when confronted with plausible-sounding anti-Android/anti-Google propaganda: - Consider the source. Is there a track record of reliability? Or the opposite? What happened the last time this source made a pronouncement? Did it turn out to be true?
- Consider agendas. Are there discoverable links to Android competitors? Does the person have clear connections to Microsoft? Google is your friend.
- Then, before you print, ask the folks who actually understand the GPL and how it works to tell you if there is any substance to the claims.
So when anyone provides an opinion about the GPL, whether a lawyer or just the usual anti-Google windbags, ask yourself, does this person know what he is talking about? I think it's happened now often enough, we should be able to figure it out, if we follow the above steps. The lawyer who wrote about the header files, Edward Naughton, has a bio page at his law firm. Today it looks like this. But he recently changed it, stripping out the name Microsoft from his bio and replacing it with "a Fortune 50 software company". Now why might that happen? Here's how it used to look, as you can verify in Google cache for a little while longer (just go to Google, search for http://www.brownrudnick.com/bio/bio.asp?ID=512&ForwdName=Edward+J.+Naughton and then click on cache), in the part where he lists his successes for clients:
- Co-counsel defending Microsoft against a putative consumer class action alleging that it had violated wiretapping statutes and common law privacy rights by designing Windows to permit third parties to place cookies on computers. Obtained dismissal of complaint....
- Represented Microsoft in several dozen lawsuits against resellers and corporate end-users of counterfeit, infringing, and unlicensed software.
Why so modest? Might it be possible that he or the firm decided it would be prudent to remove that connecting link once his article claiming Google might be in legal trouble over the GPL appeared on Huffington Post?I'll let you be the judge. But shades of Get the Facts, y'all.
Update: For the record, a friend downloaded the kernel source code today and looked at the COPYING file (and you can do it too, if you'd like to confirm it for yourself) and it says in part:
NOTE! This copyright does *not* cover user programs that use kernel
services by normal system calls - this is merely considered normal use
of the kernel, and does *not* fall under the heading of "derived work".
Also note that the GPL below is copyrighted by the Free Software
Foundation, but the instance of code that it refers to (the Linux
kernel) is copyrighted by me and others who actually wrote it.
Also note that the only valid version of the GPL as far as the kernel
is concerned is _this_ particular version of the license (ie v2, not
v2.2 or v3.x or whatever), unless explicitly otherwise stated.
Linus Torvalds
The rest of the file is the GNU GENERAL PUBLIC LICENSE Version 2, June 1991
Update 2: Steven J. Vaughan-Nichols has more:
Sean Hogle, a technology attorney, agrees. Hogle wrote, “The most objectionable aspect to the Mueller and Naughton blog entries are the wildly exaggerated claims that Android applications will be forced to be licensed under the terms of the GPL in open source code form.”
Hogle continued, “With all due respect, I don’t believe that developers are taking any risks, let alone significant risks, in this context, and I don’t believe there’s any possibility that Angry Birds will have to be GPL’d against their will. Application developers have done nothing to subject themselves to the copyleft obligations of the GPL. Linking to kernel APIs [application programming interfaces] directly wouldn’t create a derived work according to the stewards of the Linux kernel.”
And Jon Brodkin has now posted a new article, incorporating the information that Linus [and pretty much everyone in the FOSS world] is calling the claims bogus.
|
|
Authored by: Cypher3c on Tuesday, March 22 2011 @ 02:39 AM EDT |
Whta --> What. [ Reply to This | # ]
|
|
Authored by: Cypher3c on Tuesday, March 22 2011 @ 02:40 AM EDT |
Please make links clicky and put the title of the pick in your post title. [ Reply to This | # ]
|
|
Authored by: Cypher3c on Tuesday, March 22 2011 @ 02:41 AM EDT |
On-topic posters will be required to recite SCO's filings from memory. [ Reply to This | # ]
|
|
Authored by: Cypher3c on Tuesday, March 22 2011 @ 02:42 AM EDT |
Hahahaha! I have again achieved the quadrafecta.
Send in your Comes transcripts.[ Reply to This | # ]
|
|
Authored by: Anonymous on Tuesday, March 22 2011 @ 02:56 AM EDT |
Alternative to using the google cache (which will probably not last for
long):
Link[ Reply to This | # ]
|
|
Authored by: Cypher3c on Tuesday, March 22 2011 @ 03:16 AM EDT |
These claims are fairly transparent when you get down to it.
After all, why would some outside lawyer even care whether Google violated the
GPL? Unless they work for Google or someone who plans on suing them, what
business of it is theirs?
(Groklaw is a big exception, as we try to stay informed on FOSS and FOSS
opponents, so it is in our interest to analyze this stuff).
I guess you always have to look for an agenda, whether it is clickbaiting, FUD,
etc.
[ Reply to This | # ]
|
|
Authored by: rhdunn on Tuesday, March 22 2011 @ 03:33 AM EDT |
This (along with the patent suits against downstream vendors of Android) looks
like part of a multi-pronged attack against Android to hurt its
popularity/sales.
Microsoft do not have a successful product in the market (Windows Phone 7 is
struggling to get off the ground) and are late to the party. So they are doing
what they do best: bully.[ Reply to This | # ]
|
|
Authored by: Anonymous on Tuesday, March 22 2011 @ 03:35 AM EDT |
I like the double meaning in that line of the article. [ Reply to This | # ]
|
|
Authored by: Anonymous on Tuesday, March 22 2011 @ 05:00 AM EDT |
Like David Howard of the Ethisphere Business
Ethics Committee?
Who in his other job is
Microsoft's Corporate Vice President (how many of these do they have!?) and
Deputy General Counsel, Litigation Group. [ Reply to This | # ]
|
|
Authored by: Bystander on Tuesday, March 22 2011 @ 05:28 AM EDT |
It's not helping things that so many people, even journalists that are
nominally friendly to FOSS, still get important details wrong in reporting this
story. For instance, Brian Proffitt's article had the following
passage:
In "pure" Linux distros, glibc is used for this
interface job and even though it
is licensed under the GPL, application
developers can use it to their hearts'
content without fear of having their own
works fall under the GPL because the
Linux kernel developers have said (many
times) that they don't regard glibc as
a derived work, so deep interfacing with
it will not mean software using it will
in turn have to be licensed under the
GPL.
He means well, but the explanation is factually lacking.
Glibc is not
licensed under the GPL, it's licensed under the LGPL. It's the
specific terms of
the LGPL that allow applications written to use glibc to be
licensed under
other licenses according to the wishes of the application
developers. The
kernel developers, Linus in particular, have said that any user
space
application, including glibc, is not considered a derived work of the
kernel
through use of normal system call interfaces. This only makes sense,
since by
definition every C library such as glibc must interface with the
kernel through
such defined interfaces. If using such interfaces constituted
creating a derived
work, then every C library interfacing with a Linux kernel
would have to be
licensed under the terms of the GPL, and every user land
application that
interfaced with a C library such as glibc would be a derived
work of that
library and would also have to be licensed under the terms of the
GPL. That
has never been the case in practice, and has never been the stated
intention
of either the Linux kernel developers or the GNU glibc
developers.
What almost everyone writing about the story seems to miss
is that there
are at least three separate software components (OS kernel, C
library,
application code) involved in creating any complete C-language
application;
and the licensing conditions for each component have to be
considered
distinctly for each case. The Linux kernel is licensed under the
GPLv2, and this
license covers the kernel header files (to the extent that
material in the
header files is copyrightable and requires a license), GNU
glibc is distributed
under an LGPL license, which differs from the GPL
primarily in that
applications using glibc as a self-contained library are not
required to be
licensed under the same license terms. So right away we see that
it is not
necessary for the kernel code and an associated C library to be
licensed under
the same terms. Applications created to run with glibc are also
explicitly
allowed by the terms of glibc's LGPL license to use any licensing
terms desired
by the application authors. Google's Bionic C library
implementation merely
serves the same purpose for Android as glibc serves for
many Linux
distributions, and follows the same licensing conventions. As a
user-space
software component, Bionic is allowed to adopt a different license
than that
used for the kernel; and applications written to use Bionic are
allowed to
adopt a different license than that used for the C
library.
Another misconception that is frequently portrayed in media
reports is
that Google has attempted to change the license on some Linux code
from
GPL to the Apache license. Proffitt's article repeats this mistaken claim,
without providing any clarification about why this is wrong. He's not alone in
this regard.
Naughton raised serious concerns in his report
that what Google has done
with Bionic (essentially creating a library analogous
to glibc for Android) and
then licensing Bionic under the Apache license
violates the GPL. If that's the
case, he argued, then Bionic could be
considered a derived work under the
GPL and therefore any application that
touched Bionic would also be subject
to the GPL.
The problem
is that Google did not really try to change the license on
code taken from
Linux. If you look at the text block appended by Google to
header files that
were automatically processed with their scripts, you can see
where the common
perception is wrong. [emphasis added by me]
*** This header
was automatically generated from a Linux kernel
header
*** of the same
name, to make information necessary for userspace to
*** call into the
kernel available to libc. It contains only constants,
*** structures, and
macros generated from the original header, and
thus,
*** contains no
copyrightable information.
Google is claiming that the
resulting files from their cleaning process
contain no copyrightable
information. If this is the case, there is no need for
any license granting
rights under copyright law, and applying another license
to the result would be
pointless.
So what does this mean? If Google is completely correct, then
no
additional license of any kind is needed to utilize the remaining contents
of
the cleaned header files. But even if Google is not completely correct, and
there remain some pieces of the original header files that deserve copyright
protection, the license terms for those pieces remain unchanged from what
they
were, which is the GPLv2. As was explained above, having GPLv2 licensed
kernel
header files is no impediment for either a C library or any other user
space
application to be created under a different license.
As a practical
matter, whether the cleaned kernel header files are truly
free of any
copyrightable material or if they are found to contain some small
parts that
are found to be copyrightable, it makes almost no difference. Every
vendor who
deploys Android will do so with a license to use the full set of
Linux kernel
headers under the GPL, since the Linux kernel is a major
component of Android.
If they follow the terms of the GPL for the full kernel,
they will by default
also be following the terms of the GPL for the entire
subset of the Linux
kernel headers that are used with the Bionic library to
create user space
applications. No matter how much or how little of the Linux
kernel header
material is actually copyrightable, any part found in Bionic will
be properly
covered under the terms of the GPL for the full Linux kernel.
I know
that's not as simple a story to tell as taking shortcuts with the
facts to
create stories with more controversial headlines, but there should be
at least
some journalists around who still aim for accuracy in the information
they
present. Not being a professional writer, I realize my own explanation is
likely not particularly succinct or well written. Maybe others with more
literary
skill can help translate the ideas into a form that is more clear and
understandable so that more people in the general population can
understand
what's the true story here.
--bystander1313 [ Reply to This | # ]
|
|
Authored by: kenryan on Tuesday, March 22 2011 @ 07:03 AM EDT |
You'd think that the FUD campaign of Android by competitors would itself be a
sufficiently interesting topic for a journalist to write about.
I suppose though that it would mean a journalist would need to do actual
research to trace down connections, interests, finances, etc.
Offhand I can think of only one journalist willing to do that much actual work
...
---
ken
(speaking only for myself, IANAL)[ Reply to This | # ]
|
|
Authored by: Anonymous on Tuesday, March 22 2011 @ 07:30 AM EDT |
> This copyright does *not* cover user programs that use
kernel services by normal system calls - this is merely
considered normal use of the kernel, and does *not* fall
under the heading of "derived work".
But I can see why Microsoft would love such an
interpretation.
After all, imagine if any program using Windows system calls
would somehow be subject to Microsoft's control by claiming
that in itself creates a "derived work".
__
magicmulder[ Reply to This | # ]
|
- Viral - Authored by: Anonymous on Tuesday, March 22 2011 @ 08:15 AM EDT
|
Authored by: Anonymous on Tuesday, March 22 2011 @ 08:26 AM EDT |
When i read:
Does the person have clear connections to Microsoft? Google is your friend.[ Reply to This | # ]
|
|
Authored by: msfisher on Tuesday, March 22 2011 @ 08:32 AM EDT |
You should just stop the sentence there. Your advice works for all journalists
in all situations.[ Reply to This | # ]
|
|
Authored by: tz on Tuesday, March 22 2011 @ 08:33 AM EDT |
I think it would work both ways.
http://www.groklaw.net/articlebasic.php?story=20031222161942627
Here is the letter from SCO warning recipients of alleged copyright violations.
I must tell you that the list of files has everyone I am hearing from falling on
the floor laughing. We will be issuing a statement explaining why as soon as
they recover.
[ Reply to This | # ]
|
|
Authored by: Anonymous on Tuesday, March 22 2011 @ 09:44 AM EDT |
Florian Müller is a One-Man-SCO! [ Reply to This | # ]
|
|
Authored by: Anonymous on Tuesday, March 22 2011 @ 11:37 AM EDT |
Can't help but wonder.... if the kernel header files are non-copyrightable, why
do they have the GPL notice at the top?[ Reply to This | # ]
|
|
Authored by: Anonymous on Tuesday, March 22 2011 @ 11:52 AM EDT |
What's crazy is that Copyright Law
specifically identifies who has standing to bring a copyright infringement
Lawsuit.
(b) The legal or beneficial owner of an exclusive right
under a copyright is entitled, subject to the requirements of section 411, to
institute an action for any infringement of that particular right committed
while he or she is the owner of it.
Yet those .... journalists (and
I use the term loosely) that are helping to spread the FUD appear to be not
doing the one thing they should be doing:
Follow up with the Copyright
owner(s) and see if the owner(s) think Copyright Infringement has
occurred.
It's also highly amusing because it leads me to think:
This
is all MS has left? Making claims of copyright infringement for a
product to which they don't have copyright standing?
True... they have
their patent claims... at least for a while till the FSF and company gather
sufficient prior art to request a re-examination and the patents
disappear.
But is MS really reduced to making claims of copyright
infringement where the owner(s) make clear no such act has
occurred?
Where MS has absolutely no Legal standing and therefore their
opinion holds as much Legal Water as a wax-coated, wires 1 micron-thick, sieve
with each hole 1 centimeter square in size?
Wow. The last act before the
heavy, spiralling plunge? If it is, this wasn't MS going out with a roar but a
wimper.
RAS[ Reply to This | # ]
|
|
Authored by: Anonymous on Tuesday, March 22 2011 @ 12:49 PM EDT |
Even if their works are not technically "derivative", Google is still
playing the same game as Oracle, Red Hat and others of making a mint off the
free works of Linus and others, and refusing to GPL their works in the same
manner.
Technically they are right. Morally - Google is about on par with Apple, which
makes vast piles of money off of software built on a foundation of a community
of volunteer BSD coders.
By all rights, Linus Torvalds should be the wealthiest man in the world right
now (by far). Fortunately, he seems to be humble and focused enough to avoid the
problems of extreme wealth. Unfortunately, he is surrounded on all sides by the
Rumpelstiltskins of the world - Google, Red Hat, Oracle, etc - gleefully
spinning his code into gold for their own fat wallets.[ Reply to This | # ]
|
|
Authored by: DCFusor on Tuesday, March 22 2011 @ 01:26 PM EDT |
You are being too kind PJ. Here, that is a true statement.
Elsewhere, what passes for...heck, lets call it "the publishing
business" has very little to do with journalism, and a lot more to do with
publishing press releases with maybe a few added words by the "author"
who otherwise repeats a press release verbatim.
Gotta keep them advertising bucks flowing, journalism is the first to go if it
messes with that.
And it costs money on its own besides -- takes time to check even google, much
less do any real digging on your own for things not already so easy to find
(google) that anyone should know them without even having to go to google (or
any other search) specifically, because they've merely been keeping up to date
all along..
There's very little actual journalism left in the world; why do you think this
site is so popular?
---
Why guess, when you can know? Measure it![ Reply to This | # ]
|
|
Authored by: ailuromancy on Tuesday, March 22 2011 @ 01:30 PM EDT |
Newspapers and TV stations will only buy a story
once. They will buy the
first one that arrives.
If you are a journalist about to cover an event,
first
you write the story, then you go to the event.
You rush out before the end,
change a sentence or two,
then sell your story. If you wait until after the
event, someone else will have sold
the story before you.
Your story
must start with some variation of
"The end of the world is nigh!". For example,
you
could track down some crackpots to tell you what
could happen when two
high energy beams of protons
crash into each other in a super conducting super
collider. Then you can go to CERN on the day that
they will do their first
test with only one beam, and
breathlessly count down to the
apocalypse.
Many journalists do know better. Rob Enderle
said he
deliberately writes flame bait. All the
page hits and complaints make his site
look like a
good place for adverts.
The biggest cause of illness related
to Chernobil
was stress cause by the over-the-top reporting.
Clearly, all
nuclear power stations should be shut
down because
Fuk
ushima
has shown the nuclear industry
has not done anything to combat this
stress.
[ Reply to This | # ]
|
|
Authored by: ghopper on Tuesday, March 22 2011 @ 02:34 PM EDT |
May of the structures in the headers are required by POSIX and other standards.
Some of these structs were posted
to mailing lists in the early days of
computer science, and were shared widely enough (e.g. in textbook examples)
that they are essentially public domain.
So a work that has a few lines in
common with a GPL released product is not automatically covered by the GPL.
Although the compilation as a whole is a GPL work, the individual lines and
procedures can be individually
licensed and used, inasmuch as the original
source can be determined, and small excerpts may fall under fair use
even when
the original source is lost. There is no case law that clearly defines a
derivative work. (It's a
complicated situation, and I am not offering legal
advice.)
I certainly do not advocate copying without attribution, or any
other form of plagiarism. However, I am concerned
that developers will fail to
study and learn from the abundant repositories of source code, for fear that
even
looking at GPL sources will somehow contaminate their ideas and inject a
viral-GPL requirement into everything
they write. In a way, this fear is
similar to the willfull infringement claims that can apply after reading
patents. The situation is ridiculous. If the purpose of sharing information
is to promote the state of the art,
then why is there such a large disincentive
to studying the works of other people?
Artists publish books and teach
classes to promote their style. They are flattered when this becomes the
standard
that others want to copy. Is the work of engineers and scientists
really that much different?
And now I have clearly gone off topic. :)
[ Reply to This | # ]
|
|
Authored by: Anonymous on Tuesday, March 22 2011 @ 07:41 PM EDT |
As someone pointed out in a comment, Linus' opinion as to what counts as a
derivative work under US law is not definitive and may differ from that of the
courts. He isn't a lawyer (thank goodness - the world would be poorer if he
were). His expertise as to the functioning of the kernel would undoubtedly
qualify him to act as an expert witness in a case centered on this question. But
the actual decision as to what is or isn't derivative would be made semirandomly
by a judge on the basis of folksy anecdotes about baseball or bizarre analogies
to plumbing. Who can predict what a judge would actually decide if this question
was ever before them.
This statement isn't important because Linus knows more about what is and isn't
derivative than other people. Rather his statement is important because it would
provide a really solid estopple defence should anyone associated with the linux
kernel ever change their minds and try to sue for this kind of thing in future.
That is a pretty good guarantee that it wont happen.[ Reply to This | # ]
|
|
Authored by: rebentisch on Monday, March 28 2011 @ 06:59 AM EDT |
<blockquote>You thought if Florian Mueller or some lawyer who has worked
for Microsoft pitched an anti-Google/Android claim of doom, it must be so this
time? How many times will this happen before journalists stop repeating the FUD
and ask the people who know, as Proffitt just did? It's what journalism is
about.</blockquote>
Journalism does not work like this. I don't think any customer is afraid of
using Android.[ Reply to This | # ]
|
|
|
|
|