Slashdot Mirror
When Should File Formats Be Placed in the Public Domain?
wccwcc writes "A lot has been said about file formats and standards creating network effects and huge profits. That said, is there a time when file formats should enter the public domain, or is it ok for companies to sit on them forever. These are some ideas on when and how file formats should enter the public domain, just like trademarks do when they become "generic"."
Does Software Need a Special Form of Intellectual Property Protection?
Introduction
It is no secret that exclusivity is critical for extreme profitability, we call this monopoly. The founding fathers understood the tension between granting monopoly to encourage innovation but placing limits on monopoly to prevent deterrences to future innovation. These grants are often formally constructed in the form of patents and copyright, or in the form of legislation with respect to anti-trust enforcement.
The existing laws concerning grants of monopoly are becoming inappropriate or inappropriately enforced when it comes to intellectual property, especially in this digital age of perfect copies. What kind of exclusivity should be granted now and how does one maintain exclusivity as it pertains to the goal of fostering innovation.
There are many cases today where certain forms of intellectual property confer incredible network effects. A cursory analysis reveals that exclusive possession of a dominant standard of interchange confers definite market advantage and market monopoly. Whereas markets with publicly available specifications are very competitive. A standard is a form of intellectual property, and the claim is that the efforts and expenditures needed to produce that format are costly, and without some form of exclusivity innovation would be deterred. This is a plausible claim, but the question to ask is what are the limits of that exclusivity, and are there occasions where that exclusivity can be revoked. The rest of this essay explores the different types of IP protection as it applies to software, the duality between expression and utility that software possesses and suggests that software file formats may be placed into the public domain in the same way that trademarks are.
A Survey of Intellectual Property Protection
United States Intellectual Property law as specified in the constitution has always acknowledged a grant of exclusivity in order to foster innovation. However, that grant of exclusivity has always involved some trade off to encourage further innovation. With patents, the trade off has been that the patent owner possesses exclusive license to their invention in trade for fully disclosing enough information such that a skilled practitioner would be able to replicate the utility covered by the patent. Patents also place limits on the length of time the patent owner receives exclusive license. Lastly, patents primarily cover utility or functionality. There are special types of patents known as design patents that cover the appearance of an invention, not the function. Design patents provide a shorter length of time of exclusivity.
The other major form of intellectual property right is copyright, which protects expression. U.S. Law grants the owner of a copyright, exclusive rights to determine means and frequency of duplication for a period of 80 years after the death of the creator. This grant of exclusivity is received in exchange for allowing recipients of copyrighted material fair use rights. Fair use rights include rights to study, quote for review, criticism, scholarship and research. Copyright clearly does not cover utility, even though the material expressed may have utility, an example being a cookbook.
There is another form of protection known as Trade Secret protection, which basically states, if you take reasonable steps to protect the knowledge in question, you retain unlimited exclusivity through secrecy. As long as reasonable steps are taken to protect secrecy, the exclusivity is protected. If for reasons of negligence the trade secret is revealed all rights are lost.
As we see, all forms of exclusivity granted are not unlimited and all come with conditions. Additionally, the protection afforded is mutually exclusive, you cannot possess both patent protection, copyright protection and trade secret status at the same time for the same piece of intellectual property. However a piece of work may contain multiple forms of protection. So what is the correct form of intellectual protection for software?
How Should Software Be Protected?
Software has traditionally been protected by a combination of copyright and licensing. Copyright pertains to how the material can be disseminated, and licensing agreements are a way of asking the user to revoke the rights of fair use. As the author is not a legal scholar, it is not clear whether that revokation is binding since one cannot cede certain rights away. The ramification then is if copyright applies to software, does the user have the right to study and reverse engineer the application for scholarly or criticism purposes. Under license no, but under copyright protection, the user should have the right. This question is left open for further discussion.
However there are other issues with treating software as copyrightable material and that is how software possesses properties of both being expression and utility. The distinction between source code and object code has been covered extensively because of the open source movement, and both have been treated as roughly equivalent. A more detailed look reveals that source code is clearly expression, it is understandable by people but not by computers. Object code is the reverse, understandable by computers but not by people. In addition, object code when executing has utility in many cases and is not expression. Given that object code clearly may possess utility and in any normal sense of the word does not possess expression relative to people. Does object code appropriately copyrightable?
Next consider the product of object code, or running applications such as a word processor or spreadsheet. It is generally accepted that a document created on a word processor can be copyrighted. This is very clear when the document is published in a human readable physical format such as paper. But what is the copyright owner status of the actual word processing file. One part of the data file is the expression of the author, the surrounding information describing how the document should be printed, who the author is, etc. is the expression of the application and arguably the expression of the programmer. But is that information about information what is commonly called metadata, expression, and if so, is it copyrighted expression? Probably not, but then what is the IP status of the file?
The argument can be made that the file format possesses utility, describing the representation of how user defined data should be presented. If a file format has utility then it is an invention, and the proper protection would be patent, requiring sufficient disclosure that a skilled practioner could replicate it. If the file format is copyrighted, who owns the copyright of documents distributed in the word processing format? As you can see, there is not an obvious resolution.
RIght now, most file formats most likely fall under the category of trade secret, this is especially the case with current office productivity applications and desktop publishing software. Limited disclosure by the creator perhaps, but in most cases not sufficient for a skilled practitioner to replicate. Adobe's PDF is a notable exception in that full specification is published and available royalty free. It is hard to argue that the vendors of these applications have taken efforts to hide the secret description since documents using the specification have been distributed widely. It is not like the formula for Coke that is locked in a safe somewhere. What active steps software publishers have taken is by encoding the metadata in object code and even sometimes encoding the author's expression as well. This object encoding makes reverse engineering very difficult. Trade secret status is also more compatible with licensing agreements.
So if competitive software markets for applications requiring data exchange is dependent on a fully published specification being available, but the presence of such a specification implies loss of exclusivity and hence a loss of incentive to innovate, what is one to do? Lack of innovation means that consumers will suffer, but lack of competition means that consumers have fewer choices and may in some cases pay higher rates. Is there a way out?
Perhaps there is, consider another area of intellectual property law that may provide some guidance. Have there been cases, where an owner's right to exclusivity has been revoked and placed in the public domain? The answer is yes and that is in the area of trademark law. What is discussed next is analogous and is meant to open discussion not saying that software is equivalent to a brand name.
Brand names are a form intellectual property that are developed and confer some market advantage such as an association with certain attributes or functionality. Any business school student will tell you that brand creation is a costly and difficult endeavor, similar to the creation of most IP, in fact it is a form of IP. But trademarks have an interesting property in that if a brand name comes into common usage, the trademark can fall into the public domain. This has happened in the U.S. with "asprin," the brand name for acetylsalycilic acid developed by Bayer. In Canada, you go to the drug store and request the popular analgesic by the generic name unless specifically requesting the Bayer Asprin brand. In the U.S., asprin is a generic term.
If a computing standard falls into similarly wide usage, then maybe a similar entry into the public domain might occur in the name of the public interest. It preserves the initial right of exclusivity and to benefit from that exclusivity without creating a situation where benefit can occur indefinitely.
There are certain civil liberty issues associated with unilateral revocation that some may find unsatisfying. So another alternative is offered. Accepting the premise that standards are a special form of utility and not expression, create a special class of patents for interchange standards of all types of software. Standard patents would have shorter life spans, would require full escrowed disclosure at time of filing -- meaning the specification would be disclosed to the patent office, but not to others. During the duration of this patent, the owner would receive the full benefits of exclusivity for the life of the patent, except in the special case when the standard becomes so ubiquitous that it in effect has attained common use similar to what is experienced by trademarks. A measure of generic status could be some level market and usage ubiquity similar to what is used in determining when a trademark becomes deemed generic. It would only apply to applications that are involved in interoperability.
When generic use is declared, the specification comes out of escrow and is made publicly available. At that point, if the duration of the patent has not expired, competitors could license at reasonable cost usage of the specification, akin to the requirements of compulsory licensing that current patent law requires. When the patent expires, the specification falls into generic status similar to medications and is in the public domain. Changes in kind to the specification are registered under escrow but do not extend the lifetime of the standard patent. Changes in type would permit extension. The requirement that the specification be sufficient for any skilled practitioner to replicate would mean that depending on the complexity of the standard, it would range from a description to actual source code.
The creation of a special patent type is not novel, as the existence of design patents vs. utility patents demonstrates. It's scope is limited to interchange patents and would not apply to games, conversion applications, etc. It preserves the incentive to innovate, while preserving the tradeoffs that U.S. IP law maintains in exchange for exclusivity. It also supports the creation of competitive markets, by establishing the requirements for a competitive landscape.
Such a patent may also foster innovation by removing the temptation for incumbent companies to preserve vendor lock situations over coming up with new innovations making the situation analogous to the pharmaceutical industry where keeping the drug development pipeline full is crucial since patent protection expires on a regular basis. Though the recent patent grant, to Clarinex a new form of Claritin, suggests that the pharmaceutical industry is mimicking the software industry instead of the other way around. If we start getting marketed Viagra XP or Claritin 2002 we will definitely know that that is the case.
So the answer to the innovation question may rest on acknowledging the special properties of software both possessing utility and expression in different contexts. The proposals above suggest ways where specifications could be made public yet still allow companies to benefit from a limited grant of exclusivity for their efforts. The proposals are in line with protections of existing IP and in the spirit of fostering further innovation that the constitution intended.
Conclusions
If the purpose of anti-trust remedy is to restore competition to markets, a form of intellectual protection needs to be created that fosters innovation in the early stages, yet benefits the public when it becomes ubiquitous and in common usage. Trademark law offers prior precedent where intellectual property can be forced into the public domain. However such drastic action may seem punitive after the fact. Perhaps instead, a new class of patents should be created to accommodate software or standards of interchange. Patents provide the benefit of exclusivity in exchange for disclosure to foster future innovation.
Information markets are new and the special conditions they impose are just being understood. New types of protection are created to handle changes in technology as witnessed by plant patents and design patents. What is proposed is a codification and clarification to return true innovation to the software market place.
In the end this is not an economic question but a societal question as to what type of software markets we wish to have and what tradeoffs and limits we are willing to accept to create those markets.
Acknowledgements
Thanks to David Franson for reviewing an early draft and offering helpful comments and criticisms.
I think it's helpful to make this distinction because the formats can easily be made public domain, whereas the codecs are usually subject to patents and/or require licensing.
File interchange becomes much simpler if you stop worrying about codecs, and concentrate on supporting formats - Linux tools for video could read & write the AVI format (just interleaved RIFF chunks) but either use a free codec, or none at all, without creating yet another video file standard.
MS Word is a different problem - although RTF is the preferred interchange format anyway. Certainly my partner (who occasionally works as a freelance proof-reader) uses RTF extensively with her clients instead of Word native format. The RTF v1.6 spec. is available in the MSDN, and includes sample reader & writer source code.
Jon.
In early versions of PKZIP for DOS, there was a file named "DEDICATE.DOC" (actually a plain text file, not a Word document; MS Word hadn't usurped that extension in those days). The contents of this file for the version 1.1 release follows:
Dedication
----------
The file format of the files created by these programs, which file format
is original with the first release of this software, is hereby dedicated to
the public domain. Further, the filename extension of ".ZIP", first used in
connection with data compression software on the first release of this
software, is also hereby dedicated to the public domain, with the fervent
and sincere hope that it will not be attempted to be appropriated by anyone
else for their exclusive use, but rather that it will be used to refer to
data compression and librarying software in general, of a class or type
which creates files having a format generally compatible with this
software.
Of course, the fact that the file format (and application note) for this program was made freely available and usable to all, allowed numerous other companies to make their own ZIP compatible programs, many of them financially more successful than PKZIP. But that's another story...
- Mike
Microsoft obtained a patent on the ASF (Active Stream Format) audio file format on March 21, 2000. The patent is entitled: "Active stream format for holding multiple media streams."
Here is the abstract:
An active stream format is defined and adopted for a logical structure that encapsulates multiple data streams. The data streams may be of different media. The data of the data streams is partitioned into packets that are suitable for transmission over a transport medium. The packets may include error correcting information. The packets may also include clock licenses for dictating the advancement of a clock when the data streams are rendered. The format of ASF facilitates flexibility and choice of packet size and in specifying maximum bit rate at which data may be rendered. Error concealment strategies may be employed in the packetization of data to distribute portions of samples to multiple packets. Property information may be replicated and stored in separate packets to enhance its error tolerance. The format facilitates dynamic definition of media types and the packetization of data in such dynamically defined data types within the format.
Microsoft forced Avery Lee, a free software developer to revoke support for the ASF format from his "VirtualDub" GPL-ed video editing/capture program. Here is Mr. Lee's description of his encounter with Microsoft (from Advogato article, referenced below):
"Today I received a polite phone call from a fellow at Microsoft who works in the Windows Media group. He informed me that Microsoft has intellectual property rights on the ASF format and told me that, although I had reverse engineered it, the implementation was still illegal since it infringed on Microsoft patents."
Linux World News aptly commented on the effect of file-format patents on the future of free software (article referenced below):
"Microsoft is claiming patent protection on a file format. It is not at all hard to see what could happen with an extension of that claim. Any serious word processor in the commercial world has to be able to cope, somehow, with the numerous variations of Microsoft's Word format. If that format were to come under patent protection, programs like StarOffice, ApplixWare, and WordPerfect could lose the ability to work with Word files. That would not bode well for their future market share.
Intellectual property laws increasingly look like the tool of choice for those who wish to fight against free software. The ability to patent file formats, if it stands up, adds greatly to the power of this weapon. This is a worrisome development indeed."
The very idea of "forcibly placing" ones work into the "public domain by law" is quite distasteful.
Um, that's exactly what a patent is, and does. You document your method in detail for all to see, and in exchange, you can decide who uses the method for a limited amount of time. After the patent expires, potentially everyone benefits from your documentation freely.
[
The idea here is that we do not want to dictate the software used, but we want to ensure that whatever software is used, OTHER software is free to interoperate with it, including open source software. By levelling the playing field through open file formats, open source software would actually be able to compete without all the belly aching you hear about "not being able to read and write Office formats" and the such.
Bravo. Your summary of my implied position was excellent.
And I will add one more thought:
By requiring open file formats, our government would be much less likely to find itself with files that are unreadable in ten years.
If this is true, then couldn't it be argued that [DeCSS] is a valid reverse engineering exercise for purposes of interoperability with the Linux operating system?
The difference in the DVD-file-format case is that DeCSS was first released as a Windows executable, useful for nothing but copying DVDs. The DeCSS developers should have finished the Linux kernel's UDF drivers before publicly releasing the DeCSS application.
Will I retire or break 10K?