By Dr. Axel Metzger and Dr. Till Jaeger
The controversy over legal problems surrounding Free Software has been stirred further by two new statements. Within a few day's time both a legal opinion by Professor of Law Gerald Spindler of German University of Goettingen for the German Software Association "Verband der Softwareindustrie Deutschlands e. V.", or VSI, as well as a paper written by Paris Professor of Law Christophe Caron in the most important French law journal Dalloz (Dalloz 2003, issue no. 23, p. 1556) have been published. Mr Spindler's study is to be found at the center of interest, accompanied by widespread press coverage.
VSI's press release bears the title "Studie belegt Rechtsunsicherheiten", or "Study Confirms Legal Risks". It cuts Mr Spindler's paper short on statements critical on Linux. This is not surprising bearing in mind that VSI in the first place represents the interest of proprietary software industry. On the contrary, the paper in fact is much more incisive and minute. So we would be jumping to conclusions altogether wrong in stigmatising the author as an outright opponent of Free Software. The legal problems the author draws his attention to are already well known. Mr Spindler generally agrees to opinions published priorly by others, dissenting on only a few issues. On the whole he agrees to what is generally accepted by his fellow scholars in the field. As far as he dissents, his opinion is disputable. In particular, we have found some misunderstandings as well as a number of wrong interpretations of the meaning of software licences. We exemplify this by exploring some stipulations of the GNU General Public Licence, or GPL. A more complete analysis of Mr. Spindler's study is still to be completed. For this, we have to wait for the dissertation theses by Koglin and Schulz as well as a new edition of Jaeger & Metzger's "Open Source Software -- Rechtliche Rahmenbedingungen der Freien Software" which will discuss problems involved in detail.
First of all, it is irritating that the terms "freeware" and "Free Software" seem to be used as synonyms in the paper (p. 18). Software programs distributed as freeware in most cases do not conform to the Free Software Foundation's or the Open Source Initative's definition of Free Software. In the case of freeware the user is generally not allowed to modify the program code. Mr. Spindler's use of vocabulary is misleading and does not conform to common technical usage.
The study deals with complex projects in which a large number of software developers participate, working together simultaneously, or subsequently on writing software. We do welcome that the author propounds that each member of the project may bring an action for injunction based on copyright law against any violator, thus acting for each member of the project without having to name all of them personally (p. 26). This result is most important for the question how to seek legal protection for copyright law infringements. It conforms, however, to conventional wisdom. It is important, and "Linux-friendly", too, that according to the author he will lose copyright who infringes the obligations imposed by the GPL. This issaid in GPL's section four. According to German law this is to be seen as an "auflösende Bedingung", or a "dissolving condition" cf. § 158 sec. 2 BGB (p. 31).
It is, however, midleading that Mr Spindler's paper in a number of cases addresses a so-called "Pflicht zur unentgeltlichen Weitergabe der Open Source Software", or an "obligation to redistribute Open Source software for free". There is no such obligation whatsoever to redistribute, or to share software contained in the GPL. The GPL does impose certain obligations if a licensee redistributes software voluntarily, or if he obliges himself to redistribute software. The point is, that there is no such thing as an obligation to publish, or to republish software. On the other hand, according to section one subsection 2 of the GPL it is allowed to redistribute software for a charge provided that it is not meant to be a licence fee. Unfortunately the paper does not address the matter how to draw the distincion between a simple charge and a licence fee. So this issue remains to be dealt with in future research.
The author treats the matter of how to include Open Source lincences in contracts (under German contract law) with users. This discussion is to be found at the center of the paper. It is of pivotal importance, and the matter is dealt with in a rather problematic way. The author argues that shrink-wrap licences generally are void contracts provided that German law is applicable. This is a benefit of his paper. However, in addition to that the author elaborates on the difference between both phenomena, the 'classical' shrink-wrap licence as well as Open Source licences. With the former kind of licence consumer's rights are reduced to merely using the software, while Open Source licences provide for rights going far beyond that. With a shrink-wrap licence a consumer generally purchases the right to make use of a program that is reduced later at home when he learns of limitations to his rights from reading the licence. As opposed to that, such problems do not exist at all with GNU/Linux & Co. In this respect the paper is worked out rather carefully although it is easy to see that the author does not evade parallels to conventional software licences. The paper holds that Free Software licences become part of a contract by including a redistributed copy of the software purchased by the user. This is incorrect, however, as these licences are independent from the contract. Both the purchasing the software---no matter whether this is achieved by download, or by way of a distribution on CD, or on DVD---as well as acquiring rights from the licence form two seperate contracts which are altogether independent of one another. Regularly, there is no more contract as a typical user will not modify, copy, or redistribute the software at all. In this case according to German law an express consent on using the software is not at all required according to § 69d sec. 1 UrhG. This is why including free software licences in contracts has to be seperated from the mere purchase of software. Bearing this in mind, contrary to Mr Spindler's point of view considerably fewer problems arise with free software licences. ftware at all. In this case according to German law an express consent on using the software is not at all required according to § 69d sec. 1 UrhG. This is why including free software licences in contracts has to be seperated from the mere purchase of software. Bearing this in mind, contrary to Mr Spindler's point of view considerably fewer problems arise with free software licences.
According to the paper, including free software licences in purchase contracts does not fail due to the question discussed above, but rather due to a, say, lingual matter. A German-speaking consumer cannot be expected to understand an English-language licence. We doubt whether this opinion will prevail, particularly in German courts. For it would be a contradiction in terms if a consumer on the one hand would argue on the basis of an Open Source licence, whereas on the other hand obligations from the very licence were not to be seen as legally binding because they were written in a foreign language. It is altogether out of the way to argue that a consumer is not entitled at all to rights arising from a free software licence as he cannot read and understand it. We think, however, that the Free Software community should become more professional. It would be helpful indeed to bring about multilateral licence politics drawing on the specific problems of the law of consumer protection in Europe, bringing about licences in more than one language.
As to the range of the rights a user is entitled to by a Free Software licence the author pursues an approach altogether contrary to what generally is held in jurisprudence so far. Mr Spindler argues that the right to make software available publicly which is necessary for redistributing Free Software over the internet is not part of the GPL itself (p. 61). Furthermore he argues that the GPL does not comprise online, or cyber-law as due to international treaties that became legally binding in the U. S. in 1997 we have to distinguish between physical distribution of software, and the 'making available' of software (p. 44). However, this is not in line with the conventional rules of interpretation in jurisprudence. The Linux kernel has been published under the GPL mostly. So developers should have known about redistrubuting code over public networks even in 1991 when GPL version 2 was drafted (cf. Torvald's autobiography "Just For Fun").
Result:
VSI's announcement about the legal risks in developing and using Free Software was premature. This is why in the meantime it has provoked harsh criticism by Free Software lobbyists (cf. the FSF Europe's statement as well as LIVE's press release, both in German). The new study contains mostly long-known findings. It does not doubt that the Open Source model basically works. Of course, we may discuss problems already known, but this does not lead to the result proposed by VSI. Anyway the question is whether VSI has done its members a favour at all in publishing the paper, initiating that much publicity. "Proprietary" software companies, too, make use of Free Software more and more. So they also need legal standards to rely on. Propagating doubts concerning legal risks in the field will not reverse economic development. The software industry will have to deal with Free Software as another model rather than trying to push it out of the market.
English translation by Juergen Fenn.