11. Main open problems (left for the second and third papers)
1.1. An anecdotal introduction to the open source model of software development
Tuesday the 17th of June 2008 has been a symbolically quite significant day for the worldwide open source
(or FLOSS)10 community. On that day, the latest version of the most popular open source browser, Mozilla’s
6 Both the US and the European Patent Offices are quite generous in granting software-related patents (even though with some
significant differences, as I will discuss), however there are reasons to believe that the majority of these patents are, in fact, related to high-tech industries making use of software more than to the software industry itself. About this point, see in particular JAMES
BESSEN & MICHAEL J.MEURER, Patent Failure, (James Bessen ed., Princeton University Press. 2008), chapter 9 (and § 9.1.1 in
particular): “[C]redible evidence shows that software publishers have flourished so far despite the growth of software patents. Some preliminary studies suggest that although there is evidence of some detrimental effects of patents within the software industry, the effect of patents within the industry has not been serious to date. […]. [However, the arguments of authors who maintain that anxieties about software patents are exaggerated] seem to be largely directed at a carefully chosen straw man. The general concern is over software patents, not the software industry per se. This distinction is important because almost all software patents are obtained by firms outside the software industry. Bessen and Hunt (2007) find that the software publishing industry only obtains 5% of all software patents granted; most are obtained by firms in electronics, telecommunications and computer industries. […] Patents have little negative effect within the software publishing industry to date because there are no substantial patent thickets within the industry.” See also JAMES BESSEN & ROBERT M. HUNT, An Empirical Look at Software Patents, 16 Journal of Economics and
Management Strategy, 157--189 (2007) (or JAMES BESSEN & ROBERT M. HUNT, An Empirical Look at Software Patents, Federal
Reserve Bank of Philadelphia Working Papers 03-17/R (March, 2004)), collecting “evidence that software patents substitute for R&D at the firm level; they are associated with lower R&D intensity.”
7 Starting between 1998 and 2000 on both sides of the Atlantic, according, for instance, to MARCO RICOLFI, Is There an Antitrust Antidote Against IP Overprotection within Trips?, 10 Marq. Intell. Prop. L. Rev., 305--367 (2006), pp. 355—363 in particular. The author
mentions the following cases as the starting point of the “parallel development” of patent protection of software in the US and in the EU: State Street Bank & Trust v. Signature Fin. Servs., 149 F.3d 1368 (Fed. Cir. 1998), cert. denied, 119 S. Ct. 851 (1999); and In re Int'l Bus. Mach. Corp., 31 IIC 189 (2000).
8 I will not discuss the literature about the anti-commons in this paper. For further references, see in particular: MICHAEL A.
HELLER, The Tragedy of the Anticommons: Property in the Transition from Marx to Markets, 111 Harvard Law Review, 621--687 (1998); MICHAEL A.HELLER & REBECCA S.EISENBERG, Can Patents Deter Innovation? The Anticommons in Biomedical Research, 280 Science, 698--701 (1998); J. BUCHANAN & Y. YOON, Symmetric Tragedies: Commons and Anticommons Property, 43 Journal of Law and
Economics, 1-13 (2000). For further references, see the third paper of the dissertation at hand and the following works: BEN
DEPOORTER & FRANCESCO PARISI, The Market for Intellectual Property: The Case of Complementary Oligopoly, in The Economics of
Copyright: Developments in Research and Analysis, (W. Gordon & R. Watt eds., 2003); FRANCESCO PARISI, et al., Simultaneous and
Sequential Anticommons, 17 European Journal of Law and Economics, 175--190 (2004); GIUSEPPE DARI-MATTIACCI & FRANCESCO
PARISI, Substituting Complements, 2 Journal of Competition Law and Economics, 333--347 (2006).
9 See the first paper of the dissertation, § 9.2.1 for some examples of cross-licensing to solve the cross-veto power coming from
pools of software patents. Obviously, this kind of agreements, are not only apt to increase barriers to entry; they are also unable to eliminate costs arising from patent trolls, essentially patenting software solutions, without really applying them to the development of their own programs.
10 The acronym FLOSS stands for Free/Libre Open Source Software and wants to highlight that open source software is not only
(and not necessarily) free in monetary terms, but it is “free” as in “freedom”. In fact, the Frech work “libre” eliminates the ambiguity of the English word free, meaning also gratis (i.e. “for free”). See also http://en.wikipedia.org/wiki/FLOSS. I will frequently use the term FLOSS, because – as described by the Wikipedia – it is “an inclusive term generally synonymous with both free software and open source software which describe similar development models, but with differing cultures and philosophies. 'Free software' focuses on the philosophical freedoms it gives to users and 'open source' focuses on the perceived strengths of its peer-to-peer development model.”
Firefox 3.0, was downloaded by 8,002,530 people (or, at least, by as many IP addresses) and Guinness confirmed to Mozilla that Firefox effectively set a new world record as the “most downloaded software in a single day”11. This news was reported by several major media, frequently along with a much more significant
datum concerning the market share of Firefox in the browser market, according to some estimates having passed 19% worldwide during June 200812 (and being around 30% in Europe).
Firefox 3.0 is an interesting example of how powerful an open source community can be in competing with established incumbents. In January 1998, at the end of the so-called “browser war”, preceding and accompanying the famous US Microsoft (III) antitrust case, Netscape (before being bought by AOL) started an open source project, called Mozilla13 (using a peculiar license14 and keeping to itself the possibility of
releasing proprietary versions of Netscape). About 10 years later, on March 1st, 2008, support for Netscape browser from AOL officially ended. But the “browser war” had been won by Microsoft several years earlier, and that victory was sanctioned (and, in a way, “paid”) by Microsoft with the settlement it signed, on May 2003, with AOL, which had previously filed suit for damages (on the basis of the finding of facts in the Microsoft III antitrust case).15 About at the same time, the Mozilla project announced that it was going to
focus its attention on two projects, the main one being Mozilla Firefox (and the other Thunderbird, an open source mail client). In fact, not only the end of the browser war between Netscape and Microsoft’s Internet Explorer (IE) coincides – in some way – with the beginning of a “second browser war” between IE and Firefox, but the very source code of Netscape, passing through the Mozilla suite, is in some way an ancestor of Firefox’s code (even thought, in the meantime, it has likely changed enough as to maintain very pale similarities).16
For the purposes of the paper at hand, the point of this story is that – in order to compete with big software houses, in markets that they consider as strategic and in which they are willing to invest a lot of resources (and maybe risk antitrust liability) – it is frequently useful, (if not necessary), to ask the help of the open source community. Indeed, a very similar story could be told about Sun Microsystems’ Star Office, the source code of which is behind the success of the famous OpenOffice suite, competing with Microsoft Office.
I will elaborate more on this point later (in § 7). For the moment, what I want to argue is just that open source projects are sometimes better able to compete with proprietary ones, thanks to their different structure of incentives (and lack of traditional financial budget constraints). Moreover, I submit that the success of Firefox and other open source software, frequently running (also) on Windows, is lowering the switching costs potentially associated with leaving the dominant Windows operating system. And that is true because these pieces of software are available for the majority of other platforms (Linux, Mac, etc.), hence they make it much easier to migrate one’s data and settings (or even to share them running two or more platforms on the same computer, as on the PC on which this paper was written).
Discussing these switching costs, the same day as Firefox’s 3 launch something else happened that is likely to influence even more the level of the barriers to entry protecting Window’s sales. Something that is especially relevant for the paper at hand. In fact, on June 17th, 2008, the team of Wine’s developers released
version 1.0 of Wine. Building this piece of software to a stable version took 15 years of development. But what is Wine, why is it relevant, and why will we probably heard more about it? Wine allows Linux and other Unix-like operating systems users to install and run applications developed for various versions of Microsoft Windows (notice that this implies that also users of recent Mac systems can use Wine). As an illustration, I wrote part of this paper on a copy of Microsoft Word installed on a Linux Kubuntu system, needing no more
11 See Maggie Shiels, Firefox download record official, Technology reporter, BBC News, Silicon Valley, 2008/07/03 (available on
http://news.bbc.co.uk/go/pr/fr/-/1/hi/technology/7486668.stm; last visited July 13, 2008).
12 Statistics provided by Net Applications (http://marketshare.hitslink.com/) for the month of June 2008. The trend reported by
Net Applications is constantly increasing for Firefox in the last 2 years. In July, 2006 Internet Explorer had 83.57% of the global market (considering all operating systems), Firefox 11.34% and (Apple’s) Safari 3.18%. In June 2008, Internet Explorer is reported at 73.01%, Firefox at 19.03%, Safari at 6.31%.
13 For a short history of Netscape, see http://en.wikipedia.org/wiki/Netscape (last visited July 27, 2008). 14 The “Netscape Public License” (NPL), available at http://www.mozilla.org/MPL/NPL-1.0.html.
15 With Microsoft III I refer to the antitrust case, which risked leading to the dismembering of Microsoft, following J. Jackson’s
ruling 87 F.Supp.2d 30 (D.D.C., 2000). In appeal, the case was vacated and remanded, with ruling 253 F.3d 34 (C.A.D.C., 2001). The case has been ended by the Consent Decree ratified by J. Kollar-Kotelly, with ruling 231 F.Supp.2d 144 (D.D.C., 2002). Consider also that with J. Jackson’s findings of fact I refer to ruling 84 F.Supp.2d (D.D.C., 1999).
16 It is probably not by chance that Firefox’s original name (changed because of trademark issues) was Phoenix, hinting to the fact
than half an hour to complete the installation of both the latest available version of Wine17 and an original
copy of Microsoft Office 2003. Similarly, no customization or special effort was needed to install a piece of econometric software that economists frequently use: Stata 9 for Windows.18
The relevance of Wine and similar projects (that I will discuss in greater detail later on) is that they could significantly lower the so-called “application barrier to entry” that protects the leading PC operating system, i.e. Microsoft’s Windows. In fact, having the possibility of installing copies of one’s favorite (or must-have, for any reason) software on Linux (or other Unix “dialects”, FreeBSD, Mac OS X and Solaris) evidently decrease the total switching cost of changing operating system. That is true for various reasons: not only does one not have to buy a new license for a Linux version (if existing) of a software that one already owns; but also perfect compatibility of saved data is assured (and that would not always be the case using open source alternative – or even different versions of the same commercial package – for some kinds of data, including – for instance – word processor files with very complex formatting).
To keep things in perspective, notice that, for Microsoft and similar incumbents, the threat coming from similar projects is significant, but not necessarily deadly, in particular if they keep innovating at a fast pace. In fact, one could reasonably suspect that the fact that Wine was recently able to reach a stable version is also related to the slowdown in the pace of innovation for computer operating systems. In particular, the penultimate version of Windows, Windows XP, has probably shown the highest longevity in the history of Microsoft’s operating systems, having being released in October 2001 and still being the most widespread operating system, with a market share above 70% as of June 200819 (in fact, Microsoft recently committed to
support Windows XP until 201420). Moreover, it must be noted that the compatibility between Unix-like
systems and applications developed for Windows is still far from perfect. In fact, perfect compatibility (and stability of applications used outside their “natural environment”) is a very complex goal, frequently incompletely reached even by new versions of the same operating system (several computer users, for instance, are aware of more than one problem in running on Windows Vista applications developed for Windows XP21). And, to be sure, for the moment there are many more problems in trying to use on Linux an
application developed mainly for Windows XP than there are problems in doing so on Windows Vista, hence the application barrier to entry did not disappear, and will probably not disappear in the following decades; it has just been significantly lowered, so that less and less users will find it prohibitive, as long as they have other reasons to shift from one operating system to another.
What this anecdotal introduction aimed at showing is simply that open source software is, by now, a credible competitor for incumbent software houses in several markets. Moreover, it is important to be aware of the fact that an open source approach is adopted with increasing frequency by commercial software houses, precisely in those cases in which they decide to fight (with apparently dim chances of success) against established incumbents. Unfortunately, the fact that open source development may be one of the most effective tools to inject competition into especially concentrated software markets should worry us, since – as I will show – some peculiarities of the legislation on software reverse engineering could put this specific model of software development at a disadvantage in achieving interoperability through software decompilation.
17 At the time of writing these lines, the latest available version of Wine was Wine 1.1.1, released July 11, 2008. The official
website (http://www.winehq.org/) claims that this version fixed bugs preventing the working of Photoshop CS3 and Office 2007 installers (I could not verify this claim or the working of the installed software, since I do not own a copy of these recent “best seller” programs).
18 On the Wine Applications database (http://appdb.winehq.org/), Stata 9 for Windows is reported to be randomly unstable
under older versions of Wine. I did not test this thoroughly, but this is largely irrelevant, since there is an official version of Stata for Linux.
19 According to Net Applications’ “Operating System Market Share”, Windows XP has a 71.20% market share, as of June 2008,
while Windows Vista at 16.14%. The total share of Windows operating systems is 90.89% (See http://marketshare.hitslink.com/report.aspx?qprid=10 for updated statistics. Last visited July 27, 2008).
20 See http://support.microsoft.com/lifecycle/?LN=en-gb&C2=1173 (last visited July, 27, 2008). See also Paul McDougall, Microsoft Pledges Windows XP Support Through 2014, InformationWeek, June 24, 2008 (available at http://www.informationweek.com/news/windows/operatingsystems/showArticle.jhtml?articleID=208800494. Last visited, July 27, 2008).
21 For a brief summary of the various opinion concerning Windows Vista and its compatibility problems, see – for instance –
Paul McDougall, Microsoft Calls Forrester’s Windows Vista Report ‘Schizophrenic’, InformationWeek, July 28, 2008 (available at http://www.informationweek.com/news/windows/operatingsystems/showArticle.jhtml?articleID=209602050 ; last visited July 28, 2008).