In comparing free with commercial software applications contrasts are most apparent to me in four areas (excluding cost): features, innovation, usability, and stability/polish.
Free software applications tend to be limited in their feature set, frequently seeming to be a knock-off of the commercial equivalent of the application minus a generation or two. This relates directly to the area of innovation – free software applications can be wholly and quite distinctly radical in their approach – for example the free/indie game movement – but are likely the product of one person or a very small team and so frequently serve a narrow-niche problem; they tend not to fill general gaps like “word processing” or “spread-sheets” because these areas are well-served by larger, more general (and generic). These larger projects tend to lack innovation, trying instead to duplicate commercial software functionality for the free market.
Free software products tend to lack some usability and polish because the resources in the group rarely exist to support these efforts and the adhocracies that build them are challenged to integrate testing on a “finished” product. The developer/user community may enable longer quasi-beta states after a launch – relying on a volunteer network – refinements and patches flow in when and if the community creates them. Software can feel like the work-in-progress it frequently is.
An exception to may of these rules is the suite of “free” tools made available by Google. They are robust, rich in features, polished, incredibly user-friendly, and among the most innovative products on the market. However, to consider them “free” in the same sense that GNU/Linux or Open Office is free misses the point. Google is ad serving, a business model whose chief product is audience. Practically every service they offer is involved in the effort of aggregating audience or data-mining them to provide better targeting and market segmentation. Google is getting paid for their products and richly so.
One key strength of the open source community is security. Because the communities and the source code are open to inspection, exploits are found, reported, and fixed quite readily. With closed, commercial applications there can be fewer eyes reviewing code for exploits, and reports of the vulnerability to the company may go unreported to the public until the company has a patch. There have been examples where well-meaning citizens have reported vulnerabilities and eventually felt compelled by lack of action on behalf of the company to go public with warning of a possible exploit and force the company’s hand. Microsoft has been placed in this position – with their near-monopoly user-base they are the choice for black hat targeting. Patching their systems can be daunting with source code in the multimillions of lines.
Now – all that said, I compared Open Office 3.0 to MS Office 2007 to a surprising revelation: Open Office is every bit as polished as Office. Now, much of the interface metaphor in OO3 is comparable to that of MSO2003 so the observation that free software tends to be a generation behind still holds. And I haven’t found anything truly innovative in the implementation, but then I didn’t expect to, given that this is general use, mass-market targeted office productivity software. But as features and polish are concerned, this is first-rate software. I was able to ramble my way through to a document with title page, pagination, columns, images and tables with captions in a way that was not consistent with Microsoft, but was consistent with the windows computing environment and so revealed either iterative testing and development or a savant on the dev team.
I like it.
