The Web makes software a process, not a product
One of the most disruptive aspects of the Internet is that it makes all content cheap and disposable. Though various industries--from music to software--have resisted the Web's commodity urge, none have managed to escape it. Whether music, journalism, or software, the Internet makes distribution and replication cheap which, in turn, makes content somewhat transitory and, hence, less valuable in itself.
As iTunes, Google, and Red Hat indicate, the best business models for the Internet age are those that focus on services around content, rather than on monetizing content directly.
Speaking of software, in particular, we've reached the end of an era that treated software as a packaged product. Software is a process, and so demands that it be monetized through subscriptions or other service fees. We spent decades pretending that digital goods like software are the same as physical goods like tables or televisions, wrapping digital goods in copyright and patents in an attempt to make them feel like permanent products, but it's increasingly clear that digital goods really are different.
Enterprises don't buy software, install it, and run it. They license software, heavily (or lightly) customize it, run it, then upgrade/update it, and customize further. Software never really reaches stasis within an enterprise deployment. It's in a perpetual mode of change.
This is why open source has emerged and done so well: it treats software as a process and prices on a subscription basis. Most open-source models charge customers for support, updates, or other ongoing services, including access to proprietary extensions or add-ons.
In this way open source embraces the Web, rather than fighting it, and makes software development and delivery an ongoing process, fitting it to how enterprises actually consume software. Novell's SUSE Linux Studio groks this, enabling ongoing customization of Linux distributions. So does Red Hat's RHEL distribution, which lets customers subscribe to ongoing, updated software. So, too, does Zimbra, which adds to the subscription commercial extensions.
Not that open source has a lock on subscriptions. Just look at the Software-as-a-Service world, which includes Salesforce.com and its ilk that make software applications available via the Web on a subscription basis, but also includes Google, which treats software and other content as means to sell advertising, a micro-subscription of sorts.
Microsoft, of course, has been the biggest winner in 20th-century software, because it has been phenomenally successful in productizing software, making it hard for the company to adapt and embrace software's 21st-century imperative: subscriptions. But adapt it must if it wants to embrace the Web rather than be bowled over by it.
The same applies to anyone that wants to build a software business in the Internet age. If your revenue depends upon selling packaged software, rather than access to a more fluid process, you might as well fill out your bankruptcy filing along with your articles of incorporation.
Follow me on Twitter at mjasay.
Matt Asay brings a decade of in-the-trenches open-source business and legal experience to The Open Road, with an emphasis on emerging open-source business strategies and opportunities. Matt is vice president of business development at Alfresco, a company that develops open-source software for content management. He is a member of the CNET Blog Network and is not an employee of CNET. Disclosure. You can follow Matt on Twitter @mjasay. 
I do support open source software, but I think that the dooms day prophesy of the end of proprietary software is totally unfounded. Why you ask? Lets take an open source system like OpenOffice for example. What happens if I modify the source of OpenOffice? I create a fork in the code making it two separate systems. I then cannot upgrade to that latest version of OpenOffice, because my custom changes are incompatible. This is why we have APIs which allow you to hook into software without changing it, thus your stuff will be compatible with future versions. So how important is the source code to developers whom have no intention of change it? Not very, which is why open source is just better in theory.
Perhaps you're missing something.