Akari Sano is a first year law student at Osgoode Hall and is taking the Legal Values: Challenges in Intellectual Property course.
Open Source Software, the next vendor lock in?
A recent CBC article stated the federal government is seeking tenders on information about free software for the first time. It quoted Open Source software (OSS) will be considered. Two issues arose from this article for me. The article cited various reasons why the government has only started to consider OSS now. One reason for the delay, that licensees are restricted in their choice of complementary software, no longer applies given the recent advances within the Open Source community. Non- interoperability should not be a barrier anymore. The other concern is the government is seeking information on OSS, assuming it is free. But is OSS really free?
Vendor lock-in?
The article is correct in that most traditional proprietary software is incompatible with other brands. One of the reasons OSS has not been incorporated into some environments is due to lack of interoperability with programs already in place. In addition, the licensee would not be able to use other brands as part of the license agreement. This is a way for the manufacturer to ensure that their products are continually purchased for upgrades and maintenance. Essentially proprietary programs have shut OSS out, and locked in the licensee.
Microsoft released its Interoperability Principles on Feb. 21, 2008. Principle III: Data Portability is particularly interesting : “Once customers use one software product to store their data, they should be able to subsequently access that data in a form that permits its use in other software products.” Microsoft will also use a program similar to OSS called Open Format, which will be available to developers without requiring a license or royalty or other fees. Microsoft writes this will foster continued improvement and facility for third party product development. For OSS proponents, this means programmers have permission to create compatible programs. This is a significant commitment from Microsoft to Open Source. If the OSS community uses this opportunity to its fullest, it will essentially negate the vendor lock in argument. OSS programs can complement already established systems in organizations. Infiltration into the government is now possible.
How free is free?
The other barrier to the widespread use of OSS has been the cost of the program itself. This may be the bigger reason limiting its appeal to organizations such as the government and the public sector. When OSS is downloaded, it is not a nicely packaged program such as Microsoft Office. Instead it is a source code that is unique, needs customization for the particular organization’s use, and requires multiple upgrades just for maintenance. OSS requires strong technical skills on a regular basis and that costs money. In addition, the more a program becomes customized, it begins to decrease the wider community’s ability, and motivation to collaborate on it. Why would people work on a source code so unique that it is inapplicable to anyone else but that particular user? Thus the more tailored a program becomes, reliance on a dedicated computer support system becomes greater.
OSS’s fundamental concepts of interoperability, not charging royalties, and collaboration with the wider Open Source community can be questioned. This begs the question, are you now not locked into OSS? Furthermore, is OSS any different or better than the traditional proprietary software?
5 Responses
The idea that OSS cannot be easily installed and used right ‘out of the box’ is outdated and completely false. As many readers will know from experience, downloading and installing the firefox web browser is no different from downloading and installing any commercial product. This is the same for almost all open source products.
Yes and no. Firefox is a bit of a funny example because its proprietary competition (Microsoft Internet Explorer) is also distributed for free. I think a more instructive example would be comparing Windows with Linux. Windows of course has been the dominant Operating System (OS) for over a decade, and Microsoft charges for it big time ($250+ for some versions of Vista). Consider the fact that every computer needs an OS and we’re talking a lot of money, which could be saved by switching to a free open source OS like Linux.
But is it really viable for an organization like the federal government to switch to Linux on their regular workstation PCs? Linux developers claim that Linux can be everything Windows is (and more), and Dell has even started offering Ubuntu on some of their notebooks in place of Windows. Yet hidden costs abound. The software is tricky to install for an average user. Regular programs that people use all the time like Office, Outlook and Photoshop are unavailable, only found in their open source counterparts. In my experience, even user-friendly distributions like Ubuntu feel cold and complicated and require using the dreaded console (think typing lines of code into a text box like in the DOS days). Windows for all its failings has the nice quality that you don’t need to know how it works to be productive with it. Linux remains an OS built by geeks for geeks (and I say that with all respect since I consider myself one). This, I think, is indicative of a larger problem with open source software: It’s built for advanced users and not the average users who just need their computers to get some work done. As it exists now, using open source software will be a headache for users and the lost productivity and frustration will offset the savings.
I think classifying software’s usability with its license is an unfair generalization. There are many types of “open” software, just as there are many types of proprietary software. A widely used application such as Firefox provides a good example that shouldn’t be discounted by the existence of Internet Explorer. Internet Explorer being an application that I’d be reluctant to equate with Firefox in the first place, as it is only “free” in so far as you can’t see the source code, distribute it, modify it, and need a Microsoft Windows license for it to run.
While on an individual level I see the appeal of a system I can use without understanding, I’m not sure it holds up to an economic analysis, especially in office settings. In a workplace, not only can I not understand all the mechanisms of proprietary software I run, but neither can my IT person, nor can anyone other than the proprietor. An office is then tied to a proprietary vendor and their continued support or imposed upgrades (a major example of this being the Microsoft XP to Vista switch). There may very well be migration costs associated with switching to open source software, both tangible and associated with the productivity of adjusting users. However, these costs are short term, help businesses remove themselves from a coercive system, and, as government studies have shown, are not even all that high (http://www.tbs-sct.gc.ca/fap-paf/oss-ll/foss-llo/model-eng.asp).
Akari, with respect, I have a difficult time seeing how an open source license would – in itself – create vendor lock-in. I can imagine a de facto lock-in where, for example, the open source product or file format is so obscure that only one vendor supports it. But surely this could also be a problem with closed source software, and could only be exacerbated by proprietary licensing.
You raise a good point, though: software should be compared on its merits, and there may be hidden costs to some open source software. But I think there are also benefits to OSS that the government might be overlooking in lumping OSS with free commercial software (the RFI calls them “no charge licensed software”), and by viewing itself as a mere “consumer” of the software. As the CBC article mentions, OSS creates a “free market” for support and modifications. Additionally, the public benefits from OSS insofar as they gain access to the source code, which individuals can both learn from and use. This is especially relevant since the software would be purchased with public funds.
So I think we agree that a unified standard of evaluation is ideal, and that software should be evaluated on a case-by-case basis, but I do not see vendor lock-in as an inherent issue with OSS. Rather, I would argue that OSS generally creates a “free market” for support and modifications while enriching the public good.
The government software procurement process should be modernized to allow OSS to compete for contracts. So long as a given software program can fulfill the specified requirements, I see no compelling reason to discriminate simply based on source.
While you may be right that people would not want to “work on a source code so unique that it is inapplicable to anyone else but that particular user”, I suspect that a software solution designed for the Canadian government would be germane to the governments of Estonia, Colombia, or others. The work on a source code by a programmer in any of those countries could be transferable to related or identical systems of other countries.
Furthermore, OSS adoption by a government could spur the development of a novel industry in that country, designed to cater to the OSS maintenance and upgrade needs of their own government and OSS-adopting governments of other countries. Whether such a model would be cost-effective remains to be seen, but the nature of OSS appears uniquely suited to keeping such costs low. The question governments, including our own, should be asking is not what they risk by adopting OSS, but rather what they risk by being late to the party.
Comments are closed.