HP tosses WebOS out of frying pan into the open-source fire

Hewlett-Packard's decision to release WebOS as open-source software doesn't bode well for the future of the project.

There are two common outcomes when companies convert a complicated proprietary project into open-source software. One is that a vibrant community of contributors grows up around the project, expanding its abilities, broadening its popularity, and making it into a better component of a broader technology package.

The other is that the project, tossed over its sponsor's transom, sinks beneath the waves.

I think HP would like the first outcome based on Chief Executive Meg Whitman's high hopes: "By contributing this innovation, HP unleashes the creativity of the open-source community to advance a new generation of applications and devices." But I expect the second is more likely.

HP tried shopping WebOS around, but finding the options unappealing, bet on a less conventional direction. But here's the big cautionary tale that leads me to my pessimism: Symbian.

That operating system began as a proprietary project, but Nokia and its backers, presumably seeing the sustained success of the Linux project, took it in the open-source direction. It was too little, too late, though. Developers weren't interested enough in building Symbian.

So even if HP tries to keep WebOS as an actively developed operating system for tablets, for example, it may be open-source in name only. It wouldn't be the first time somebody threw a party and nobody came.

WebOS didn't have a glorious future in any case. HP acquired WebOS with its $1.2 billion Palm purchase last year, but short-lived CEO Leo Apotheker decided to ditch WebOS and the mobile devices it powered. When Whitman took over, she decided to review the company's options, and HP announced the plan today after weeks of deliberation.

HP's promise to keep developers actively working on WebOS sounds more like the plan for the Linux-based MeeGo OS that Nokia pushed aside in favor of Microsoft's Windows Phone: The operating system now merely has an "opportunity to significantly improve applications and Web services for the next generation of devices." That's not the kind of statement that tells developers they need to divert any of the attention already devoted to iOS, Android, and possibly Windows Phone, Windows 8, and Blackberry 10.

There are other open-sourcing failures, too. Sun Microsystems switched StarOffice into the open-source OpenOffice.org, but the code base was messy enough and the developers controlling enough that outsiders were repelled. The project muddled along for years, prodding Microsoft Office only as far as opening up its file formats. It was only when Oracle acquired Sun and ran roughshod over the project that some developers cared enough to fork the code base and try to strike off on their own with LibreOffice.

After that, Oracle washed its hands of the affair and foisted it on the Apache Software Project, which oversees many open-source projects. IBM remains interested, but OpenOffice.org doesn't have bright prospects. Sun's Java and Solaris open-source projects were equally fraught.

AOL got farther when it created Mozilla to open-source Netscape's once-proprietary browser. But it took more than half a decade of work before Firefox clawed its way to relevance, and that only happened with a complacent Microsoft that parked Internet Explorer.

It's possible WebOS could fuel some organic, grass-roots mobile OS project--especially given WebOS' Web-like app programming model--but the mobile OS teams at Apple, Google, and Microsoft are about as far from dormant as is possible right now.

Why ever would a company want to release a once-valuable asset as open-source software that anyone can use for free? Or for that matter, contribute resources to an open-source project at all?

There can be good reasons. One of them is undermining your competitors. If you can build a vibrant community around a free product that your competitor charges for, you can steal a little of that competitor's thunder.

Yahoo, for example, supports the Apache open-source Hadoop project that competes with Google's in-house equivalent, MapReduce, for analyzing mammoth data sets. Yahoo may not have vanquished Google, but Hadoop is catching on widely and reducing Google's competitive advantage by making it easier for rivals to match its abilities. Even Microsoft is embracing Hadoop.

So might open-source WebOS exert some pressure on the incumbent mobile operating systems? I don't think it likely.

That's because the key to relevance in the OS world is apps. An operating system still can be useful to power slot machines, electronic billboards, and factory-floor robots. Old-school phones had all they needed--a dialer, address book, and text-messaging interface. But if you want your OS to shape the future of mobile devices, you need an OS that lets people play games and post status updates. WebOS, open-source or not, lacks that support.

Note that there's already an open-source mobile OS out there today that has plenty of apps: Android.

After Google releases Android versions' source code, academics, Amazon, CyanogenMod programmers, or third-tier device makers sink their teeth into it and build their own versions of the operating system.

It's true that it's not a terribly communal effort, though. Only after Google is finished planning and developing Android internally does it release the source code. Android 3.x, aka Honeycomb, never even made it to the open-source state. Google isn't violating any licenses, but it's not showing much interest in sharing control over Android.

Perhaps WebOS will shame Google into letting some others into the Android party, especially if an open-source WebOS becomes a fruitful proving ground for new mobile technology that shows Google that Android could benefit from a broader perspective.

I wouldn't bet on it, though. And even if it does, that's pretty small consolation for HP.