BeOS: Free as in Beer

Free R5 Distribution, Stinger Strategy Promise Mindshare, Money

Scot Hacker, 1/25/00

The BeOS user and developer communities are busy digesting the implications of two major announcements made by Be on January 18, both of which bear profound ramifications for the popular and corporate acceptance of BeOS. First, the company announced that release 5 (R5) of BeOS will be freely downloadable and distributable for all non-commercial purposes. On the same day, Be announced an official shift in emphasis from the desktop to the appliance market.

The two press releases almost seemed to work at cross purposes to one another. If Be is re-orienting themselves to target the burgeoning appliance market, should users and developers fear for the future of BeOS on the desktop? And if so, why push R5 out to the world so aggressively? I don't think these two strategies are contradictory, and here's why: There are potentially immense revenue streams locked away in these initiatives -- revenues which could guarantee rapid growth for both the BeOS desktop and appliance markets. Rising tides raise all ships, and each strategy has the potential to do great things for the other.


Free BeOS represents more than a naive attempt to capitalize on the popularity of free software. While it's true that the world is coming to regard bits as "too cheap to meter," the most important thing about free BeOS distribution is the reach it allows. BeOS is suddenly unchained from its pedestal and allowed to wander the Internet unfettered. It now becomes possible for publishers to attach BeOS CDs to magazines, for software libraries and FTP sites to distribute BeOS legally, for users to install the OS onto as many machines as they can find or to pass out BeOS CDs to strangers on the street. BeOS software vendors can give the operating system away along with their apps, in an attempt to (as Gobe says) "get users involved with BeOS" when they try their products.

Be has been talking for a long time about viral marketing, and this is it. The power of the meme is about to kick in, as the operating system begins to spread like the common cold. This strategy will enable us to reach hundreds of thousands, or even millions of new users quickly and cheaply.

Be is going to take a huge revenue hit on this, up front at least. But while investors seem to see negative numbers in the prospect, I see it as a simple marketing expense. In exchange, Be will see a level of exposure they haven't enjoyed since the days when BeOS for PowerPC was available as a free download. But with the x86 market being ten times larger than the PowerPC market, and the public now better primed to understand the concept of multi-boot computing, Be and BeOS developers are about to see an influx of new users the likes of which they've never witnessed.

People are funny. While they're more than willing to spend $500 on a copy of Photoshop, they're reluctant to plunk down $69 on an entire operating system -- one with the potential to completely change the way they think about computing. By distributing BeOS for free, Be overcomes this odd fear of spending a nominal sum for a great piece of software.

FreeBe will be absent a few royalty-bearing technologies, such as the Real Player, the Opera browser, and some codecs. However, it will still be possible to purchase the full version of BeOS in a box with manuals, for installation into a real partition. As for questions about whether BeOS or some part of it will at some point go open source, the possibility has indeed been broached by CEO Jean-Louis Gassée in at least one recent interview. Whether, how, and how far any possible open source initiative from Be will go remains to be seen. For now, BeOS will be free as in beer, not as in speech.

Barriers Trampled

The most fascinating aspect of this initiative, however, is how it's going to be accomplished technically. One of the biggest roadblocks to gaining new users is the fact that people are unaccustomed to working with disk partitions and setting up multi-boot machines. People fear that partitioning their drives will damage their existing Windows installations, or that they'll lose data. And while Be has already made this process more simple and painless than any other OS vendor in history, the perception remains.

The method by which FreeBe will be distributed will not require any partitioning at all. Instead, users will be able to download a Windows installer. When run, this installer will create a giant but normal file directly in the Windows partition (techniques will be available for Linux users as well). Into this giant file, the installer will write a BFS (Be File System) volume, and then install BeOS directly onto that. When BeOS is launched, Windows will be kicked out of memory and BeOS will boot up from that volume/file. We're not talking about emulation here; this is real BeOS, running from a real BFS volume. It just so happens that that volume is not residing in its own partition. As a result, users will not experience a performance hit of any kind, so long as their Windows filesystem is not fragmented.

Think about what this means. In one fell swoop, Be completely eliminates their three largest barriers to entry:

The potential for widespread titillation is huge. For more information on free BeOS (scheduled to be released sometime this quarter), see

And What About Stinger?

So all of that is pretty exciting, but there's another side to the coin. At the same time Be is making their largest push ever to garner new users, they've also announced that the "Internet Appliance" is to be the company's new emphasis. Be has been talking for quite a while about the burgeoning appliance market, and began showing their IPAD prototype several months ago. Since that time, the company has put significant work into Stinger -- a slimmed down version of BeOS with a highly customizable UI, designed to run in a very low footprint while still delivering high-bandwidth media performance. If you're curious as to what exactly an "appliance" can be and why I think this market is going to be huge, read "Stinger: What It Is" at BeNews.

Let's be honest with ourselves. Despite its many technical advantages, more modern native technologies, and superior ease of use / installation, BeOS has, for whatever reason, not rocketed to popularity like Linux has. Explaining that little mystery is a subject for another day, but the fact remains that BeOS has been a slow burner in the consumer space, where people purchase just one copy at a time. In the industrial / OEM space, vendors may be purchasing Stinger in orders of thousands of units at a time. Of course, OEMs will pay for their copies of Stinger, unlike FreeBe users. More importantly, competition in the appliance space is very different from what it is in the consumer space. The field is new, the requirements are different, and it just so happens that Be has an operating system nearly custom tailored to meeting the needs of that market. As I've said here before, this is where the twin vectors of low footprint and high performance meet, and BeOS is ideally suited for the job. Unsurprisingly, Be is being met with enthusiastic, open arms by appliance vendors. One of the first to sign on is the futuristic Qubit -- a wireless digital clipboard running BeOS. The night of the unveiling of TransMeta's Crusoe chip, CNBC showed a Crusoe device running BeOS -- the perfect appliance CPU meets the perfect appliance OS. Major vendors like National Semiconductor and Compaq have signed partnership deals with Be, and more will surely follow.

The Fallout

The remaining question for existing BeOS users regards Be's commitment to the desktop, and to the thousands of loyal end users and developers Be has worked so hard to win over. Will the new emphasis on Stinger mean Be puts less effort into the desktop side of things? The full answer to that question remains to be seen, and may depend on the ultimate success of FreeBe, the ensuing vitality of the developer community, and new revenue streams resulting from Stinger deals. There have been repercussions, though. For instance, the nascent Be Magazine I announced last month was cancelled when the publisher became unsure of the future of BeOS on the desktop (which is why I'm back here at Byte).

My feeling is that it's too early to be worried about such things. There are too many good prospects (read: lots of mindshare and lots of money) embedded in these announcements to worry just yet.

BeView Content Archives