Here is some information about an attempt to prepare an open source license that passes EU law muster.

Torger Kielland, a student at the University of Oslo who is writing a thesis on copyright law and the implications involving the Free Software Movement, emailed me with a possible way to circumvent the GPL. He states:

I make a modification to a GPL'ed program and construct it as a "dif" - a
file which only states the differences between the original code and the modified code and distributes this, the dif does not contain any of the original code, and would in my opinion not qualify as a "derivative work" of the original Program. Hence it should not be necessary to distribute it under the terms of the GPL.

However, this raises two questions:

1. If I distribute the dif TOGETHER with the original program (for instance in a .zip-file stored on a website or a CD-ROM), would this qualify as a distribution "as part of a whole which is a work based on the Program", according to GPL § 2 para. 2?

2. Even though the "dif" itself does not qualify as a derivative work, it enables the licensee to create a derivative work if he has access to the original program. Does this make me as a distributor liable for contributory infringement?

In my opinion the answer will be no in both accounts, but I'm looking for
inputs.

Thoughts?

Open Source Development Labs (ODSL) has announced that they are providing $4,000,000 to fund the start of the Software Freedom Law Center. The new center will be headed by Eben Moglen, with other luminaries involved, such as Lawrence Lessig, Diane Peters (General Counsel of ODSL), and Daniel Weitzner (involved with W3C). The idea is to provide free legal services to open source non-profits and developers.

Can I dynamically link software subject to the GPL, make my code proprietary, and not violate the GPL? How about the LGPL?

Wikipedia defines dynamic linking thusly:

Dynamic linking systems place the majority of the linker code in the underlying operating system, in which case it is known as a loader. At compile time the linker only writes down what libraries the executable needs and checks to make sure they are being called properly. When that program is then executed, the loader finds these libraries and links them at that point, either at loadtime or during runtime
when the library is actually referenced.

[Quotes from Wikipedia are subject to its license.]

From the GPL FAQ, the Free Software Foundation thinks so, at least respecting the GPL. Quoting from the FAQ:

If the program dynamically links plug-ins, and they make function calls to each other and share data structures, we believe they form a single program, which must be treated as an extension of both the main program and the plug-ins. This means the plug-ins must be released under the GPL or a GPL-compatible free software license, and that the terms of the GPL must be followed when those plug-ins are distributed.

What are your thoughts?

Sun Microsystems has announced that it is making 1600 patents available to the open source community.

Upon a careful reading of their announcement, it looks as if these patents are only provided to that part of the open source community that is using the Common Development and Distribution License (CDDL), which is Sun's recently OSI-approved license that they intend to use for OpenSolaris. The patents don't seem to be available for just anyone in the open source/free source communities.

There seems to be some difference of opinion whether certain kinds of "linking" creates issues under the GPL. The question mostly seems to be whether a derivative work is created or not.

Let's suppose that there is some code that is subject to the GPL or the LGPL (call it the "library") that is "statically linked" to other code. Does that make the other code a derivative work of the library?

Why is this important? If I link a GPL'd library to other code and then distribute the result, I may be forced to use the GPL.

It seems to me that we first have to create a common ground for just exactly what "static linking" means. According to Wikipedia, static linking:

"embeds a library directly into the program executable at compile time by a linker. A linker is a separate utility which takes one or more libraries and object files (which are previously generated by a compiler or an assembler) and produces an actual executable file. Internally all references to code located in the library are replaced with a pointer to the code; since it cannot move independantly of the executable, this fixed offset will never change.

One of the biggest disadvantages of static linking is that each executable ends up containing its own copy of the library. "

[Quotes from Wikipedia are subject to its license.]

Does static linking create a derivative work?

Alex Thurgood, a partner in the French IP law firm Cabinet Michel Richebourg emailed me and suggested this topic. (Thanks, Alex.)

Quoting from Alex:

One question that has always bothered me with free software licenses is their overall lack for any consideration of jurisdiction outside of the US. While as an IP lawyer, one might let pass an ironical smile knowing that the GPL was born in the US, and that this same country was one of the last major industrialized countries to sign up to the Berne Copyright Convention, it is nevertheless interesting to note that the majority, if not all, of the clauses of the GPL have been drawn up uniquely with US law in mind. The upshot of this is that, IMHO, the GPL is probably inapplicable in many countries outside the US. Indeed, in France, the GPL was considered by the scientific research community to be so inadapted to French law that it drafted a new free software license, called the CECILL.

End quote.

Germany has similar issues, as far as I can tell. For example, the analysis of the Munich decision upholding the GPL suggests that at least some Germans have a VERY different attitude toward licenses in general. (Maybe that is the reason for Alexander's antipathy toward the GPL. After all, he's from Germany.)

Australia has problems with the GPL as well. Under Section 68(1) of the Australian Trade Practices Act, disclaimers are apparently void if they don't comply with the language in the act, and it may bring down the entire agreement. See this at pages 76-80 for more detail. Australian info is courtesy of Ian Oi, Special Counsel at Blake Dawson Waldron in Australia, in his attempts at certifying another OSI license.

Maybe U.S. lawyers have been lulled to sleep by the promises in the Berne Treaty that all copyright law is the same.

What are the problems with the GPL overseas? Are they important? To whom?

IBM has announce that it will pledge 500 patents for those working in open source. The official announcement is here. Do you think that other companies will join in this activity? If so, why?

If you use and modify software subject to the GPL, but only use it "internally" as a way to provide online services to "external" third parties, have you violated the spirit of the free software movement?

Is the GPL3 likely to "plug that hole"? If so, what will happen to existing companies that use this business model?

Gumout, a thoughtful contributor to this blog, raised the following issue, and it's a good one to discuss.

[Starting gumout quote]

The GPL's section:

"2 b) You must cause any work that you distribute or
publish, that in whole or in part contains or is derived
from the Program or any part thereof, to be licensed as a
whole at no charge to all third parties under the terms of
this License."

is preempted by section 301 of the Copyright Act. In one of
the most favorable decisions ever for contract rights
concerning copyrights the Seventh Circuit in ProCD, Inc. v.
Zeidenberg; 86 F.3d 1447, 39 USPQ2d 1161 stated:

"Rights equivalent to any of the exclusive rights within the
general scope of copyright are rights established by law
rights that restrict the options of persons who are
strangers to the author. Copyright law forbids duplication,
public performance, and so on, unless the person wishing to
copy or perform the work gets permission; silence means a
ban on copying. A copyright is a right against the world.
Contracts, by contrast, generally affect only their parties;
strangers may do as they please, so contracts do not create
exclusive rights. Someone who found a copy of SelectPhoneTM
on the street would not be affected by the shrinkwrap
license though the federal copyright laws of their own
force would limit the finders ability to copy or transmit
the application program.... But whether a particular license
is generous or restrictive, a simple two-party contract is
not equivalent to any of the exclusive rights within the
general scope of copyright and therefore may be enforced."

A person finding a CD containing source code and a copy of
the GPL *is* affected by the license... They are invited to
join in the modification of the source code. The GPL is not
a simple "two party" contract. It is a recursive multi-party
contract of adhesion that attempts to publicly regulate
derivative works.

From the GPL:
"Thus, it is not the intent of this section to claim rights
or contest your rights to work written entirely by you;
rather, the intent is to exercise the right to control the
distribution of derivative or collective works based on the
Program."

Here's some of what Congress had to say concerning section
301:

"The intention of section 301 is to preempt and abolish
any rights under the common law or statutes of a State that
are equivalent to copyright and that extend to works coming
within the scope of the Federal copyright law. The
declaration of this principle in section 301 is intended to
be stated in the clearest and most unequivocal language
possible, so as to foreclose any conceivable
misinterpretation of its unqualified intention that Congress
shall act preemptively, and to avoid the development of any
vague borderline areas between State and Federal
protection."

There are few sure bets in this world but here's one:
The GPL's section 2 b) is D.O.A. in a Federal Court of
Appeals.

[ending gumout quote]

Do you agree with this? Disagree? Why?

The 12/27-1/3 edition of NetworkWorld has a two-page Microsoft add beginning on the second page. It cites a Yankee Group study that "a significant switch to Linux from Windows or Unix could cost three to four times as much without delivering tangibly better performance or business value."

I'm interested if anyone has any actual experiences in a midsized or large business setting that either supports or doesn't support this Yankee Group conclusion.

Groklaw is starting a series on the relationship between patents and antitrust. The initial discussion is here.

In the war story parts of legal seminars on open source and in my own practice, it is becoming clear that senior managers don't know that their programmers are using open source. This is an especially bad thing when its use is discovered in connection with due diligence in an equity investment, going public, or being acquired. Senior managers are routinely giving representations and warranties in contracts that the company uses no open source, and upon checking they are wrong.

What can you do to manage the use of open source in an IT organization?

Let's explore whether the GPL is a copyright misuse. Copyright misuse is a legal doctrine. It was created as an analogy to the patent misuse doctrine. In both cases, it attempts to prevent a copyright holder from exerting greater rights to stuff (a legal term of art) that isn't protected under copyright/patent law than the actual rights that exist by virtue of being a copyright/patent holder.

An example might be useful. Suppose someone develops and patents a salt spreader. That company then refuses to sell the salt spreader unless the purchaser also agrees to purchase salt from the company. In effect, to get the use of a patented product (the machine), one is required to buy a non-patented substance (salt). That, in the eyes of the law, is patent misuse.

The first well-known case, as far as I know, that applied the copyright misuse doctrine is Lasercomb America v. Reynolds, a 1990 Fourth Circuit opinion. In that case, Lasercomb created a CAD software product that made steel rule dies. Steel rule dies are used to cut and score cardboard and paper for making boxes. Lasercomb offered to license, and actually licensed to some customers, the software product. In its license agreement, it provided that the licensee agreed not to develop a competitive product for 99 years. At the time, the copyright period for the software was less than 99 years.

The Fourth Circuit held that the presence of this clause was a copyright misuse, which created a defense to a copyright infringement claim. In rendering its decision, the court asked the question whether Lasercomb was using its copyright "in a manner contrary to public policy...." In effect, the Court stated that exercising rights not available to one under copyright may be a misuse, which creates a copyright misuse defense. (Note also that the Court stated that such a defense does not invalidate the copyright. The copyright can still be exercised if the offending action is stopped.)

Let's translate that to the GPL. The GPL in essence states that a licensee is not authorized to copy or distribute code unless the licensee does certain things. Therefore, the licensor is exercising its exclusive rights (the right to copy and distribute) in creating certain conditions to copying and distribution. This is potentially similar to a named user restriction in a proprietary-type software licensing scheme.

What the GPL does may be different from a restriction that lasts longer than the period that copyright protects the code, and therefore not subject to the copyright misuse doctrine.

Let me have your thoughts.

We've added support for RSS. Unfortunately, it won't tell you when new comments are made, just new topics. Hope you enjoy.

Another company is adopting a "dual license" strategy: one set of code for a product licensed under the GPL and the other commercially licensed. This new one is for an object-oriented database program.

For Open Solaris, Sun has created a new license, the Common Development and Distribution License. It's based on the Mozilla Public License 1.1. General information on this license is here. Here's an eWeek commentary on the new license.

There is a new group in the U.K., called the Open Source Consortium. It is apparently a group of companies that get a significant percentage of their revenue off of services related to open source products. It also indicates that it will provide insurance. This insurance seems to be like a performance and payment bond, rather than indemnity insurance for open source infringement claims.

One of the most important issues related to the willingness of corporate America to adopt software licensed under the GPL -- maybe the most important issue -- is contained in Section 2 of that license. It states that if you modify GPL'd software and then distribute it, you must license the modified software "as a whole at no charge to all third parties under the terms of this license [the GPL]."

The phrase "all third parties" takes on particular importance here. For example, if "all third parties" means each and every third party in the world, whether a licensee of the modified software or not, the distribution of the modified software to a subsidiary of a company could trigger the right of a competitor to get the code. (For the purposes of this discussion, let's ignore what the remedies are for not complying with a competitor's demand. Focusing on the meaning of "all third parties" is useful without talking about remedies.)

If on the other hand, "all third parties" means only those who are licensees, directly or indirectly, of the modified software, then a distribution to a subsidiary, who would have no interest in licensing it to a competitor, creates a potentially different business result.

Let me know what you think "all third parties" means and why.

The GPL pretty clearly wants to cover "compliations" as well as derivative works. Just like the creation of a derivative work, if you create a "compliation" that includes software licensed under the GPL, the GPL wants to require you to license the compliation under the GPL. This intent is manifest in the next to last paragraph of Section 2, where it states "the intent [of section 2 of the GPL] is to exercise the right to control the distribution of derivative or collective works based on the Program."

So what's a collective work? A good example is a book of short stories by different authors. There, each author retains the copyright in the individual short stories, but there is a separate copyright to the "compilation." Another example might be the collection of music on the CD distributed with the last edition of Wired Magazine. The theory is that there are choices made about what to include and not include; the nature of those choices are expressive and therefore copyrightable.

However, the next paragraph of the GPL states, "In addition, mere aggregation of another work not based on the Program with the Program ... on a volume of storage or distribution medium does not bring the other work under the scope of this License."

It looks like the GPL states that you must make compilations subject to the GPL except when it is a compilation.

I don't think this language works. Your thoughts?

Here's news about a new version of the GPL, due next year.

If you (a U.S. company or resident) download open source code, and some of it was created by a developer in Cuba or other U.S. embargoed country, have you violated the law? Does anyone care, including the U.S. government? How would you protect yourself from this problem?

There is one thing that troubles me about Microsoft's new IP indemnity, the text of which can be found here.
The language just after the exclusions, states "You will reimburse us for any costs or damages that result from these actions." This suggests an indemnity back from the user to Microsoft, in the event one of the exclusions occur. In an end user license, this kind of language (a)is rare and (b)has none of the limits on them that Microsoft has allowed for itself in its indemnity. This language should give prospective end users pause.

Microsoft has broadened its IP indemnification to cover most of its customers, not just the big ones. The official announcement is here. Groklaw has posted a deconstruction of the text of the indemnification. Much of the textual analysis on Groklaw is IMHO inaccurate. Looks to me like Microsoft is giving a real indemnity.

However, is it truly a valuable thing that Microsoft is providing a full indemnity? First of all, few software companies, in my experience, can avoid giving a full IP indemnity to customers, so maybe the response is "welcome to the club, Microsoft."

Second, and more importantly, consider the answers to the following questions. Has Microsoft ever been sued for patent infringement related to technology in one of its products? (Hint: Yes, a number of times.) At the time of such patent infringement claim, did Microsoft have a contract that fully indemnified the users? (Hint: No) Has anyone sued a Microsoft end user for patent infringement over a technology that Microsoft put in its products? (Hint: I don't think so)

This points out a fact of software business life: a commercial software developer normally needs to fix a patent infringement claim for one of its products for business reasons regardless whether an indemnity if given to customers or not. Regardless whether an indemnity is provided, Microsoft has to fix its patent problems.

The giving of the indemnity is therefore, in my opinion, not giving much if anything away by Microsoft. A marketing ploy? Possibly.

However, the fact that there is a company that views it in its best interest to defend patent infringement claims is potentially a real distinguishing feature between MS Windows and Linux.

My two cents. Your comments welcome.

EWeek reports that Microsoft may be attempting to control certain Internet protocols.