These two essays by Richard Stallman cover a lot of ground, and I think he does a great job of laying out the issues as both ethical and empirical questions. While the ethical questions are perhaps inherently debatable, the empirical ones are not. It is neither necessary nor in our long-term interests for software to have owners.
So the question is, why do we still use commercial software? In the SLC Digital Media Lab, which I set up, we use a closed, commercial operating system (Mac OS X) as well as several software packages which have built-in restrictions on copying, use, and modification (Adobe Creative Suite, Autodesk Maya, Apple iLife and Final Cut, etc.). On the other hand, we also use free software (Audacity, Blender, Firefox, Jahshaka, NeoOffice, VLC, etc.).
At one time, free operating system software (GNU/Linux) was not up to snuff in stability and compatibility. Beating Windows in stability was not a high hurdle, but it was a quite valuable achievement, as evidenced by the commercial success of Red Hat Linux in the mid- to late 1990s. And since the advent of OpenOffice.org (formerly StarOffice, released in 2000), there has been no really valid excuse to run Microsoft’s office suite, much less its OS.
On the other hand, Apple’s Mac OS had long been the choice of creative types, beginning with the desktop publishing boom of the 1980s. In a neat progression of increasingly data-intensive media, “industry standard” spots were won by Adobe in 2D imaging (Photoshop, 1990), Digidesign in audio (ProTools, 1991), and Avid in video (Avid/1, 1994). All of these software packages ran on Mac OS, and only by the late 90s were they ported to Windows.
Part of the Mac’s appeal has always been its look and feel. By only allowing Mac OS to run on their own hardware (except for a brief licensing experiment in 1997-8), Apple was able to provide a relatively trouble-free user experience while concealing some of the inner workings – an approach that would not work for Microsoft’s exceptions-filled market.
By the late 90s, Apple was sensing the Linux zeitgeist as well, and made the savvy decision to turn to BSD as the backbone for its next-generation OS (dig into a thorough Mac OS history here). Released in 2000, OS X was a Unix variant and therefore brought many of the benefits – with the notable exception of open interoperability – of the GNU/Linux systems. This made it very easy for creative professionals to stick with Mac. For the same reasons, colleges like Sarah Lawrence find Mac to be the easy choice.
But the question involves more than technical details. Remember that, for professional users, changing OS and critical software involves abandoning not only a skill set – how to use particular software – but also shifting a whole network of practice, the social milieu that allows these professionals to learn, belong, and agree upon “industry standards” in the first place.
It is obvious from the history of industry standards that the best tool does not always win. Networks of serious users, typified by working professionals, establish their own standards and criteria as much to exclude outsiders and identify insiders as for any technical reasons. Just as in government politics, professional politics are not about the particular platform, which often changes, but rather the group identity and shared practices. In this context, any way of avoiding massive change in workflow will be preferable, even if that course is considerably more expensive.
As a learning institution, SLC is always negotiating its standing with respect to the fields we teach. Critically, as a liberal arts college, we are not training students for a vocation, and therefore graduates are not necessarily expected to know the industry standard tools. But it is just as critically important for students to understand and engage in the ongoing development of practice in a field, the ultimate arbiters of which are the working professionals. A student will not really engage with psychology until she argues the current issues as set forth by professional psychologists. When studying a digital field, she will not appreciate it fully until she engages with the current industry standard software.
When I’m not at work, I record music. I use a variety of tools, including the free software Audacity for editing. But when it comes time for tracking and mixing, the best free software available, Ardour, is just not up to the task. It can’t address all the I/O of my audio interface, the latency is too high, and it crashes. When the performance is up to snuff, I’ll make the switch. But I intend to keep up my familiarity with ProTools and other commercial software, because of the particularities of their workflows.
One of the most misunderstood aspects of the transition to a digital workflow is that, often, the new tool does not allow any new functions. Given enough time, effort, and budget, the old analog tools could be made to achieve the same effect. Analog purists, especially the well-funded variety, use this argument to refute the merits of digital. But nobody works in a situation with unlimited time, effort, and budget! Again, much more important than the function of a tool (once it has achieved relative stability and usability), to the serious practitioner, is the effectiveness of the surrounding workflow in achieving a result that is up to industry standards. Therefore the industry standard tool is much more valuable than the sum of its features.
We must not forget Stallman’s point that open and free cooperation is a more healthy practice than these exclusionary politics. But heading in that direction, we run up against powerful social realities, and deep questions about the nature of expertise. In the short term, compromises are called for.
Why don’t we use free software? The question should be, how do we get free software to become the industry standard?
2 replies on “Industry Standards and Free Software”
Three notes for the history of free software…
In the 1960s, contributed code libaries were widespread. Someone who wrote a payroll program for the IBM 1410 environment, for example, might place that code where other programers could find and use it. Some user societies worked essentially as swap meets. The Association of Small Computer Users in Education is one that started as an IBM programmers group still exists, though in evovled form: http://fits.depauw.edu/ascue/. Free software should work now and in the future because it worked very nicely in the past.
Let’s honor Kermit, Pine, and Mosaic. These were/are free software of really good quality–and were contributed to the public domain by righteous institutions (Columbia University, University of Washington, and National Center for Supercomputing Applications). Each of these helped open the Internet in ways nobody else was ready to do when these went public. Free software has been transformative at key times in the history of computing.
Actually, most of the time Tom can count to three.
The item that didn’t make it into the previous posting:
Free software does make its way into commercial, copyrighted products. Example: uPortal getting adopted as the basis for the front end to SCT’s (now Sungard) Luminis portal. So, the free/commercial distinction is not always a case of mutual exclusion.