Home -> Law Blog Directory -> Legal Journals Blogs -> The Journal of the Business Law Society

Legal Journals

: The Journal of the Business Law Society

Abandoned Software

By Derek Conrad

Piracy, the copying and distribution of software without permission from the copyright holder, is a major concern for those involved in the IT industry.  Losses due to software piracy have been estimated at 33 billion dollars per year, [1] although other groups claim these estimates are too high. [2]  Most piracy consists of downloading software for free, instead of paying retail price.  However, there is a small subset of software piracy that falls into a grey area of the copyright law.  Older software works enjoy copyright protection.  But unlike a copy of Microsoft Windows or Halo 3, in some cases it is unclear who owns the copyright on the software.  Furthermore, a lot of old software, especially video games, is unavailable at retail.  And unlike books, music, or movies, old games are stored on easily damaged, decayable media, such as floppy disk.  Given the underlying goal of copyright to deliver works into the public domain after giving the author a period of protection, there is value in preserving "abandonware".  Unfortunately, the current law of copyright does not adequately balance the two goals of intellectual property in the United States.

Several aspects of software design separate video games from other types of software.  Word processing software is functionally driven in that various packages are highly substitutable.  Video games are generally more creative works, featuring characters and situations and rulesets that differ from other games.  They are more akin to movies than to word processors, and retain cultural value in a way that function-based software cannot.  An old version of a spreadsheet competes in the market with newer spreadsheets.  Companies depend on adding new features or increasing software efficiency in order to sell their product.  Old versions of functional software thus have a direct impact upon the market for new versions, and the owner of the copyright in an old version certainly has the right to stop selling it in order to promote a newer version. 

Contrast this with old versions of creative works.  While there are often newer versions of old creative works that are available, they are not always good substitutes for the old version.  Movies are remade all the time, and multiple versions of the same work can create synergistic market effects, driving demand for both products as people who appreciated either version can seek out the other if their interest is piqued.  Older creative works can become part of the cultural heritage, and stand on their own.  Old video games do the same thing, especially amongst those who were around to see the early days of a new art form evolving into the multi-billion dollar industry it is today.

Sadly, a lot of early video games are impossibly difficult to find, and even if a retail copy can be found, they might not run except on obsolete hardware.  Unlike music or movies, which can almost seamlessly be migrated to new mediums, games require a specific hardware platform.  Old hardware supplies dwindle over time, as there is usually not a market for producing replacement parts.  (Hardware, like our spreadsheet example, is functional.)  Since these early games are still protected by copyright, it is illegal to copy them without permission.  Games like Pac-man or Asteroids are in little danger of being lost, and are still published.  But many lesser-known games are in legal limbo.  Companies that no longer exist who originally owned the property might have been sold piecemeal or just vanished.  Often, the new copyright owner doesn't know about the game, or in some cases, there is no legal owner.  In order to preserve this abandonware, I will look at three possible fixes.

First, copyright owners can release their copyright to the public domain.  This is an ideal solution, as it allows anyone who is interested to pick up and play or alter the game without fear of legal consequences.  But copyright release can be problematic.  Most obviously, if no one is certain who owns the game, no one can release the software.  There are also misconceptions about copyright law.  Some owners confuse copyright with trademark, and believe that if they release the copyright on old software, they will not be able to use or control parts of that work in the future.  It is easy to see the source of this confusion -- a trademark that is not defended will cease to become valid.  But copyright does not work the same way.  If I were to write a new version of King Lear, my version would be copyrighted even though the original is a work in the public domain.  Furthermore, software could be released with a license similar to the GPL. [3]  A copyright owner could release the software but maintain copyright on the characters and story in a game, allow copying but not sales, or any other license term they would like.  The other big problem with release by the owner is that many companies might think the software might have some future value, and hesitate to give anything away for free.  This concern is valid, but has been addressed by the Library of Congress.  [4]  A person who already owns the software is allowed to make copies and circumvent copy protection for archival purposes.  While this does not allow access by the general public to old games, it does allow old software to be stored in archives and migrated to new media if necessary, paving the way for it to be released when the copyright on the work expires.

The second way to preserve old software is to dramatically shorten the copyright period for those works.  A length of 20 years could ensure that old games are preserved and that copies are available.  This approach faces serious difficulty.  First and foremost, there is probably not enough political power to push such a change through congress.  Software companies would object, and there would probably be a lot of court activity trying to determine which software would be under a short period and which would fall under the usual regime.  Secondly, some old software does retain market value for long periods of time.  While the majority of works fade from the public view after a period of time, some works remain economically viable throughout their copyright term.  It might be seen as a serious injustice to throw the baby out with the bathwater by limiting the market potential of steadily popular works. 

The third solution is a sort of middle ground between copyright release and shortening the term of copyright.  A statute that allowed the free copying of software not available for sale would have a dual purpose.  First, publishers would have incentive to provide downloads or retail sales of old games.  Secondly, software who's copyright pedigree has been lost would be added to the public domain immediately.  The availability of software would be ensured while also preserving the value of those works which are still viable.  Many old titles are very small, so storage and bandwidth costs would be minimal, and Internet sales of software have become commonplace and inexpensive.  Holders would also be able to generate modest revenue streams for software that would be too expensive to market at retail.  I believe this approach provides a lot of flexibility for copyright holders while maintaining a good balance with the public domain. 

There are a few caveats to any system for preserving old games.  Publishers might not want to maintain technical support for old titles (which could be a nightmare given the increasing obsolescence of the original hardware platforms).  They also might feel that old titles could diminish their public image.  Rules releasing companies from support requirements or warranty can deal with these issues, without depriving the public domain of works.  It is also arguable that poor titles will fail to be seen by many people, and that even bad works are part of our cultural heritage that should not be lost.   

The controversy over abandonware is still a minor one.  But even the oldest software is still almost certainly still copyrighted, and given the exponential increase in the number of works over the last few decades, the issue of abandonware will become more important.  With the growth of the Internet and data storage techniques, we have the ability to prevent losses of important parts of our history.  Addressing the issue while the stakes are lower might be the best way to strike a balance between the public domain and copyright holder interests.

[1] Business Software Alliance report, 2006 (pdf)

[2] Ars Technica

[3] GNU Public License 3.0, Free Software Foundation

[4] 37 CFR Part 201

Full post as published by The Journal of the Business Law Society on November 28, 2007 (boomark / email).