What if Sun fails with open source?

The more I read about Sun CEO Jonathan Schwartz betting the company's future open source, the more I grow concerned that if it fails, Sun will be the harbinger of sorrow for the rest of the open source world.

Sun is arguably the most important open source vendor right now as Schwartz has bet the company on software instead of its traditional hardware revenue. (I'm not discounting Red Hat's place, just that RH has been on the open source path since the beginning.)

Sun's strategy is audacious and reshapes the way that everything is done, but it's not clear that the strategy is correct or that Sun's existing corporate structure can execute on this enormous change. Staff reductions and other cost-cutting measures have little if anything to do with the switch to open source. Those measures would need to be taken regardless as the company is simply too bloated and expensive to run even if it generates a decent amount of cash.

Sun's approach--at least the way I'm reading it from Jonathan Schwartz's statements, is about making the software totally free and trying to sell support and hardware. This clearly diminishes the value of the products and doesn't offer a mechanism that encourages people to pay for software. It also puts an unnecessary burden on the notion of open source--such that if Sun is wrong, everyone else will look wrong too. But, Sun's approach is quite different from most (all?) of the open source start-ups and also different from Red Hat, the obvious leader.

The most successful open source companies have figured out ways to encourage people to pay for software. This usually includes a commercial license that removes the open source license restrictions. Typically, we see the base "open core" product plus some type of value added feature or service set that can't be obtained for the community version. The idea of simply selling support fell by the wayside for most companies at least one year ago.

If we take away Java and Solaris, Sun's only revenue-meaningful open source product is MySQL, which is also the only ubiquitous product in the Sun portfolio. Beyond MySQL, it's hard to find a category killer or even a highly adopted open source component that they generate revenue from.

If Sun is going to be a software company then it needs people to pay them for software. Period.

Note: Let me clearly state my biases:
- I'm not looking to get flamed here. I think this is a legitimate concern. - I admire and support Sun
- Java is my language of choice
- I have several friends at MySQL
- A company I founded theoretically competes with Sun in the SOA and integration areas

[joetesta70]

If Sun fails, no one would notice anymore - same goes for Prince. 1999 was 10 years ago.

[johnnykapalua]

Time for management change at Sun.

[ITRebel]

The idea that giving away free software would somehow make a profit like selling software was a flawed idea from the start. Most people who have propagated this flawed business idea have almost religious zeal in their enthusiasm for it. Unfortunately, most of these same people are too young to remember that giving away free stuff and worrying about profits later is exactly what caused the dot com failure a few years back. If a "bait and switch" approach is now the open source community's new way to make money, this will likely also fail. Companies that use software do not mind paying for it if they know that the software provider is going to be around in the future. Unfortunately, most paying customers will be skeptical about whether the open source start-ups who are now offering value-added proprietary software will be there in the future. It now looks like the only way that open source software providers can succeed is by becoming proprietary software providers, but even this is likely to fail because of their "bait and switch" method.

[daverosenberg]

Don't confuse "free" and "open source". There is no bait and switch if users are aware of the differences in the product versions.

[ITRebel]

I do not know what the difference is between the old model of giving a watered down free trial and then upgrading to a subscription. Ultimately, you are selling proprietary software or you do not stay in business.

[Penguinisto]

@:ITRebel":

The trick is to not sell the software, but to sell the services behind it. RedHat makes a tidy profit from selling support, not the software. IBM rakes in billions of dollars per annum from doing the same thing.

You may want to do a bit of research in the future, eh?

/P

[gggg sssss]

Sun aint no IBM. Never was.

[ITRebel]

Penguinisto,

Mr. Rosenberg's blog clearly stated that selling services no longer works. You need to sell value-added software on the back end to make a profit. In other words, you need to become a proprietary closed source company. I have been around a long enough period of time to know these things without doing research. It is just common sense. It might have worked for Red Hat, but it certainly has not worked for Sun and there will be many, many more failures.

[JoeF2]

And MySQL has done that quite nicely before they were acquired by Sun. They always had two licenses: the GPL, and, if you didn't want to use the GPL, for whatever reason, you could buy a commercial license. As far as I know, MySQL was profitable.
This seems to be the path that Mr. Rosenberg advocates.

[daverosenberg]

To be clear, I didn't say you needed to become proprietary. You need to have *something*, open or closed that differentiates the products to encourage people to pay.

If you are not trying to build a business and just want to make software free then this approach is not applicable.

[daverosenberg]

JoeF2 is correct. There is no incentive for users to become customers without something beyond "just support." The dual-license model is predicated on this approach.

[ZUrlocker]

Dave,
the model of open source at the core and "value added" software in a subscription is the model being adopted at Sun. It may take some more time for product lines to figure out and deliver their value add components, but that's the approach. In effect, Sun is going to use the "MySQL model" of building a profitable software business on open source. It's not just about selling support or selling hardware. (Though of course, selling hardware can be a good business also.)

But I agree, it's important for Sun to do this right and there's lots of work required to make sure Sun can become profitable and growing once again.

--Zack

[ITRebel]

Unfortunately, for smaller companies that something beyond "just support" will most often be embedded in a trade secret that cannot be made available through open source.

[ITRebel]

One more comment. Was MySQL profitable? They had 12 million in revenues in 2004, but I do not know if they were profitable. I am really just following up on JoeF2's comment and wondering if anyone has a reference on this.

[alicia234]

I was just reading their CEO's blog, he highlights just how much money they're making on Software - looks like most of the profit and growth is there. That's not their problem -their real problem is all the high end hardware that banks depend on.

http://blogs.sun.com/jonathan/resource/LineItems.pdf

[ITRebel]

I looked at the pdf that Alicia234 references. MySQL actually showed negative billings growth in the first two quarters of this year. It is not clear at all to me that MySQL is a profit maker. I see no record of this. They were taking in venture capital funding as late as 2006. Also, could one reason that Sun is in financial difficulty be the $1 Billion that they paid for MySQL? This does remind me of the valuations of the dot com companies that were not profitable and never would be profitable.

[pfenner]

Hopefully Sun can help MySQL execute better. Marten Mickos talks a good game, but his track record in delivering quality software is horrible, vs. Sun. MySQL is a highly politicized environment though, I doubt they can turn it around that quickly, perhaps its time for Marten and team to turn it over to professionals. Too much infighting and politics can kill that business.

[gggg sssss]

mySQL has grown because it was free. If you wanted to pay for a database you could always have paid for Oracle, MS SQL Server or DB2. If you wanted something from a 2nd tier player you could always have baught Pervasive and other pices of crap. The world has no room for another costly database.