Is open-source a dead-end for desktop apps?

Despite success on the server, open source hasn't conquered the desktop

Fans of the open source desktop productivity suite breathed a sigh of relief this week, when a group of prominent developers announced they're breaking ties with Oracle and launching a new fork of the suite, to be known as LibreOffice. The new suite will be managed by an independent organization known as the Document Foundation. Whether the Document Foundation will be able to sustain LibreOffice as a significant competitor to Microsoft Office, however, remains an open question. Among the skeptics is none other than the "father of Java" himself, James Gosling. Like many other former Sun employees , Gosling left Oracle earlier this year, citing conflicts with the database giant's management style and culture. Since then he has been increasingly critical of Oracle and its handling of the various open source properties it gained through the purchase of Sun — in particular, the Java platform itself. Naturally he's a big fan of the open source model, as he explained in a recent interview for the Basement Coders podcast. "The place where it falls apart, though," Gosling says, "is for desktop software." History seems to support Gosling's view. While there are numerous high-quality open source desktop applications available, few of them — with the exception of Firefox, perhaps — have caught on with the mainstream public. Because open source applications are generally available free of charge, this raises troubling questions about the public's perception of the quality, efficacy, and value of open source desktop software. Can the open source model can really sustain desktop application development, or is open source desktop software a failed proposition? Where are the Office-killers?

There's an idea out there that Microsoft Office is overwrought bloatware, that it serves no purpose but to lock customers into the Windows world, and that the market desperately needs a replacement. I must come clean: As a die-hard Microsoft Office user myself, I don't necessarily agree. But if we accept that competition is a positive force in any market, the market for desktop productivity suites is a troubled one. Office has been the leading force in this market for a long, long time. Past competitors, including WordStar and WordPerfect, did not transition well from DOS to Windows and soon dropped by the wayside as Office became the dominant force in the industry. Today we see new competitors in the market, particularly in the form of Web-based productivity suites such as Google Docs and Zoho. Notably, however, although both of these products are available for use free of charge, neither is open source. Most of the other traditional desktop software suits that have appeared over time, such as SoftMaker Office, have been proprietary software, too. Even LibreOffice began life as proprietary, commercial software. It was originally developed as StarOffice by a company called StarDivision, which was acquired by Sun Microsystems in 1999. Sun open sourced the StarOffice code as in 2000. Even then, however, Sun supported development by offering a proprietary, commercial fork of the suite under the StarOffice brand. There are other open source competitors available, including the AbiWord word processor, the Gnu Calc spreadsheet, and the KOffice suite. Even among Linux users, however, none of these projects has a user base that can compare with that of And none of the Office competitors — whether web-based or desktop software — even remotely approaches Microsoft's installed base, which consistently hovers at upward of 90 percent of the market. You don't charge me, I won't call you

What makes it so hard to create successful desktop applications using an open source development model? One obvious problem is the business model — or lack thereof. Most of the prominent, successful open source software caters mainly to the data center and developer markets. The Apache Web server, the MySQL and PostgreSQL databases, Eclipse, Perl, Python, Ruby, and (arguably) Linux itself are just a few examples. Most of Sun's software portfolio — the bulk of which was available under an open source license — fell under the same category. Some of these projects are largely volunteer efforts — the programming languages in particular. Of those that do make money, however, most use a commercial support model. Enterprise customers who rely on open source software for mission-critical applications are willing to pay hefty license fees for the ability to gain knowledgeable product support with a simple phone call. But that model breaks down for desktop applications, where self-service is increasingly the norm. "My personal view is that if desktop software requires a support call, you have failed," Gosling says. Desktop UIs have become so standardized and online help so readily available, in fact, that few desktop software vendors even bother to ship printed manuals with their products anymore. The most obvious alternative model, perhaps, is an ad-supported one. But while advertising works to support Web-based applications to some degree, the same is less true for desktop software, where users have been taught to associate unexpected advertising with spam and malware. Sun tried using advertising to support its software, particularly by offering advertising in its installers, and the result felt instantly odious. Spare a dime for a developer?

So far, the Document Foundation has made good moves to promote LibreOffice. It has already made beta versions of the LibreOffice suite available on its website. It has announced that the enhancements made by the Go-OO team — a de facto branch of the code maintained by Novell — will become part of the main codebase immediately. And it claims that it aims to be "as developer-friendly as possible." But how friendly is friendly? "The key milestone I think will be if someday the Document Foundation can claim a headcount of developers that equals or exceeds that which Oracle has working on," Rob Weir, an ODF architect with IBM, writes on his personal blog. "In the end code talks, and developers write code." The very best developers, however, won't necessarily find the Document Foundation's stewardship of LibreOffice any more appealing than Oracle's. As Gosling says, "Look guys, I'm an engineer. I don't want to be an engineer as a hobby. I don't really care about being fabulously wealthy, but I do like to eat." The fact of the matter is that a truly viable business model that can sustain open source development of desktop software has yet to be found. And without a clear revenue stream to support LibreOffice development, the project's future — like that of its many open source cousins — remains dim.

Join the newsletter!

Error: Please check your email address.

Tags Special ID

Show Comments