Thursday, January 05, 2006

Aharonian v. Gonzales

On January 3d, Judge Marilyn Patel of the Northern District of California (of Napster fame), handed down an opinion in Aharonian v. Gonzales, a pro se case brought by Gregory Aharonian seeking declarations that (1) computer source code consists entirely of algorithms and data structures considered by him to be unprotectible ideas or processes under 17 U.S.C. 102(b); (2) that "idea" and "expression" are unconstitutionally vague in violation of the Due Process Clause of the Fifth Amendment; and (3) that software is not covered by the Copyright Act.

Aharonian's interest (and what gave him standing, the court found, rejecting the Government's challenge on that point), was his development of an automated prior search tool that will assist in finding relevant prior art so that people may avoid patent and copyright infringement. His tool is to consist of three parts: (1) an expert system containing a programmatic representation of patent and copyright law; (2) a database of existing source code; (3) an analytical tool applying (1) to (2) to determine the likelihood of infringement or invalidity.

Judge Patel was troubled by the generalized nature of the attack on all software as consisting of ideas or algorithms. Relying, interestingly, on among other sources, Wikipedia, she noted the choices available to programmers in expressing their ideas, and on that basis rejected Aharonian's across-the-board challenge, further holding that if Ahaonian "copies source code that is protected by copyright law, he infringed the copyright regardless of whether the ideas underlying the source code are patentable."

The attack on idea (and "expression") as unconstitutionally vague was rebuffed with a flourish: while the terms are inexact, "the fact that words in a statute require a certain degree of judicial interpretation is not merely permissible -- it is bedrock assumption of our common law system."

7 comments:

billb said...

Do you think that he has a fair use argument under which his unauthorized copying of source code would be allowed given his proposed scenario?

William Patry said...

I think that on a case by case basis there might well be such an argument.

billb said...

I thought you might (based on the similarity of this method to what Google is doing and your opinion thereof).

I'm curious where he's going to get this library of source code to compare against that would require him to seek such a declaratory judgement. If it's from the Open Source/Free Software community, then there's no problem because the licenses for these kinds of software allow exactly what he wants to do. If it's from the proprietary community, then either he's stealing the software (i.e. industrial espionage) or he has a license agreement with the copyright holder which allows him to posses the source.

He need not sell the tool itself to third parties, but simply offer a comparison service against his library. In such an event, he need not show the contents of the library that match submitted code, but simply tell the submitter that thier code infringes (in his estimation) so-and-so's copyright and supply a contact address. Thus he need not violate copyright in order to develop the tool in question.

Clearly trying to sell the tool (along with its library) could get him in trouble with anyone he'd licensed proprietary source from.

maz said...

I apologize for posting this under Aharonian - my "comment" was misdirected to this post - I am interested in getting your comments on the new KISS Catalog decision by Judge Fischer finding the anti-bootlegging legislation constitutional (2005 WL 3485878)

Michael M. said...

Does this ruling present a conflict of sorts with the 2nd Circuit ruling in the 2600 case? If you don't mind indulging me 'cause I ain't no lawyer and my understanding of these issues probably is insufficient to grasp all the nuances involved, I'd like some help in figuring out how to square these two decisions.

In 2600, the claim was made that prohibiting DeCSS was a 1st Amendment violation. The ruling held that source code was not speech, and therefore there was no 1st Amendment issue. Here, Judge Patel is ruling that source code is used by software designers to express ideas; futhermore, she seems to be implying that it is that expression that is protected by copyright law, even if the ideas underlying the code are not patentable.

So how is something used to express copyrightable ideas not considered speech? Paintings and musical notations are not "spoken" but they are considered speech in the 1st Amendment sense, aren't they? If Congress passed laws prohibiting use of the color red in watercolors or prohibiting use of a certain chord in compositions, wouldn't those laws be considered 1st Amendment violations? Would the reasoning be that even though a specific color or chord may not be patentable, it can still be used to express ideas, and those expressions are subject to copyright and entitled to 1st Amendment protection? If that's the case, then I don't understand how source code can be singled out as ineligible for 1st Amendment protection when source code is every bit as useful for expressing ideas as any other tool.

Max said...

/* In 2600, the claim was made that prohibiting DeCSS was a 1st Amendment violation. The ruling held that source code was not speech,
*/

Actually, I seem to remember the decision being that source code was not *always* *protected* speech. There is quite a lot of speech that is not protected. For instance, telemarketing is considered speech, but it is heavily regulated.

From the appelate opinion (http://www.eff.org/IP/Video/MPAA_DVD_cases/20011128_ny_appeal_decision.html), "Computer programs are not exempted from the category of First Amendment speech simply because their instructions require use of a computer. ... Having concluded that computer code conveying information is 'speech' within the meaning of the First Amendment, we next consider, to a limited extent, the scope of the protection that code enjoys."

Yes, there are Perl poetry and obfuscated Perl contests (I'm sure other programming langauges have similar activities, but I'm not familiar with them). A Perl poem was even published in the Economist several years ago. But that doesn't mean "for (in i = 0; i < max_val; i++)" conveys any kind of expression that needs First Amendment protection.

The rest of the opinion goes into *why* the code for DeCSS fell into a regulated speech context. Then again, placing the DeCSS code on t-shirts is a form of speech that doesn't fun afoul of the DMCS, and has a very obvious message, so it's not as regulated as crafting a program that breaks a statute.

Max said...

DMCA, not DMCS