First, what I am describing is a software lifecycle change facilitated by open source, rather than a business model. Software deployers will switch from procurement to adoption without any intervention from vendors. The question really is how vendors will remain relevant in this market, not whether it will happen.

Second, this is not a support-only model. It assumes that corporations will want the value-added content of a "subscription" (Sun's model) or "enterprise version" (RHAT and MySQL model). Value-add can include patch management, performance tuning, additional utilities and more. Governance rules like SoX will also likely make any enterprise using software for a mission-critical purpose secure a service contract and a warranty/indemnity. Finally, many businesses still fear open source licenses and so want commercial licenses for production systems.

I thus think people are more than willing to pay for what was previously free if what they are paying for reduces costs or adds value. Remember, it's only the software that's free of charge, not the people who work on it! And the whole situation is a logical consequence of software freedom, not a vendor construct.

Hi,

I do not understood from Simon's post that he was advocating a support-based model, just pointing out that with the increasing availability of software, people first build the application, then figure out if they want to pay for the platform/pieces they are building upon. We works with a lot of commercial open source vendors and see this again and again.

You mention "The dual license model (MySQL) or a gated access to products model (Fedora/RHEL) are more effective than support-driven business models", yet those two companies successes build upon an existing community and userbase. In other words, the software was adopted before it was fully commercialized. And it keeps happening: people prototype applications in a spare server based on JBoss or LAMP, or directly on top of apps like the ones we package in BitNami. Then the app gains popularity inside the organization and then the "proof of concept" gets moved into production. Thus is critical that the software is readily available and easy to install and get going with. Why is Ubuntu has a bright future in the server space? Because they have already tens of thousands of developers that are familiar with it, having adopted it as their desktop.

Best regards

Daniel

Shareware? What are you talking about? Open source and adoption-led approach puts more power into the hands of buyers. That way they can try the software and see if it works. That's a good thing.

The days of selling closed source enterprise software that for millions of dollars are gone.

--Zack

"The dual license model (MySQL) or a gated access to products model (Fedora/RHEL) are more effective than support-driven business models."

Gated access to products? I can slip the latch of RHEL just by heading to ftp://ftp.redhat.com/pub/redhat/linux/enterprise, or if I'm feeling less ambitious, to centos.org.

People pay for RHEL because they need (or at least think they need) enterprise building, updating, and support. That is precisely the model Phipps is speaking of.

The situation is quite similar with MySQL, which "includes 24x7 Production Support" with MySQL Enterprise (http://www.mysql.com/products/enterprise/support.html). It is a clear mistake to think businesses use Enterprise solely for the commercial license.

Zack,

>That way they can try the software and see if it works.

Doesn't / didn't shareware let you do the same thing?

> The days of selling closed source enterprise software that for millions of dollars are gone.

If only that were true....but don't take my word for it, check out the revenue filings from Oracle, Microsoft, SAP or IBM.

Guys,

Some thoughts on a very interesting line of thinking...

Open Source products are like pumping air into a hot air balloon - sure, you attain some amount of lift - but the more air you push into the envelope the more vents out. This venting come in the form of forking, or innovation based evolution. The amount of community usefulness will determine how fast and high the craft will rise and no amount of forced hot air will make it happen dramatically faster or push it higher. I think development tools for example have a higher inversion layer (in terms of being useful to more potential contributors) than something specialized. Simple economics - supply and demand, the smaller more specialized apps will always have a lower demand (e.g.. smaller community) and a higher price.

I like the shareware comparison - it makes sense and represents the model more accurately. What we see today is like shareware with a SAAS component. Rather than buying the shareware you pay by the feature. Want better virus protection, add this - want the ability to make waffles - add that! It's still shareware...

Even so - service models offer a greater level of control in your business. You aren't going to make a VC jump for joy with it (if that's your only path to profitability), but you'll make payroll every month.

Oracle, Microsoft, SAP and IBM - listed above, make a killing in the services space and a greater portion of these services are being delivered from India and China. (great places to be building developer awareness and more business friendly environments) I would consider a service component essential to any Open Source business model. Service organizations teach people how to make use of your product - in these days where Open Source is just as complex as any multi-tier MS based development environment people need help to learn your product. Good service programs can also jump start the community effect in your forums and motivate people to transition from end users to contributors of both ideas and code.

Still - saying that big time enterprise software is going away is like predicting the death of the mainframe. It's a safe thing to say but probably not how events will transpire. Both will (and have) continued to evolve - which I suppose is also safe to say... :)

Cheers, Arthur...