Skip to main content

Thoughts on "What Sun Should Do"

Posted by terrencebarr on November 26, 2008 at 1:36 PM PST

(Notice and Disclaimer · This is not an official statement from Sun Microsystems. Nor does it reflect any particular insider knowledge. It is simply my personal observation, opinion, and reaction to some of the current discussions around the web)

Tim Bray's piece on "What Sun Should Do" is getting a lot of attention - as well as some of the follow-ups such as on JavaLobby and RedMonk.

I think it's a good and necessary discussion - but it's easy to oversimplify and miss important points. One of my favorite sayings is: "Every complex problem can be boiled down to a solution that's simple, attractive, and easy to understand - and wrong." It's a tendency we fall into easily - and what I have been missing in the discussion so far is the focus on client-side technologies and products and the role they play in technology-based business models.

Tough Questions

At the heart of the discussion are the questions: How does Sun get back to consistent and profitable growth overall as a company? What should a successful technology and product portfolio look like? And which projects and technologies don't support that portfolio and should be cut or let go?

Some people are advocating that Sun should focus on back-end and web infrastructure and maybe cloud computing and basically ditch everything else. To me that narrow focus is very dangerous and amounts to Sun cutting off it's own legs.

Here's why:

Fighting Commoditization

I think if IT history has shown us anything it is that narrowly focusing on a particular piece of a technology environment is not enough to sustain a long-term and profitable business model. Why? Because eventually the competition catches up and turns your space into a commoditized component. Even if you are the best, the most scalable, the most secure, or the have the best price-performance. Eventually, your competitors produce something that's good enough and cheaper than what you are offering. It's a fact of technology diffusion in capitalist societies.

A technology vendor you need to constantly fight that trend for commoditization - by moving up and sideways and providing a range of technologies and products that complement and build upon your strengths. You need to get involved on multiple levels with customers and users, encourage adoption, drive mindshare, leadership, and brand awareness, anticipate the every-changing needs of your customers and bind them closer to you to establish a value-chain that you can continue to leverage and monetize at multiple points over the long term.

Why Client-Side Technology is Critical

Client-side technology and services are a critical beachhead in fighting commoditization. Look at Google, Adobe, Microsoft, Amazon - a huge part (or even all) of their business model and revenue is based on back-end technology, infrastructure, and services. But they enable and solidify that business with customer- and user-facing client-side technology: Be it end-user applications, client runtimes, developer tools, or storefronts. That is how they interact with millions of customers, drive deeper into their customer's lives, and tie users into their back-end services and product portfolio. That client-side component of the strategy is essential to their business model. They own the customer/user relationship and that gives them the unique position to deliver all the services on the back end.

Imaging Google or Amazon simply providing white-box search or product logistics services and have someone else take care of the client-facing aspect of the business. Or Microsoft without their application franchise (Office, etc). I think it's easy to see that their business would be a lot less valuable and, more importantly, a lot less unique than it currently is. In a nutshell, client-side technology and footprint is much harder to establish but also provides a much greater degree of stickiness and leverage to drive a whole range of interrelated offerings.

Sun's Differentiators

Java, NetBeans, and OpenOffice (among others) have been very successful in providing exactly that beachhead for Sun. Java has generated a lot of revenue for Sun over the years (as witnessed here), has been instrumental in closing many infrastructure deals, and is a good business on nearly 6 billion devices around the planet. NetBeans has been the vehicle for Sun to enable developers to add to the overall value of Sun's technology stack and drive the technology deeper and wider. And OpenOffice has spearheaded open source adoption and its huge popularity has created a valuable relationship and channel with users and customers around the world. And last but not least, JavaFX is Sun's entry to participate in the RIA space - an increasingly important mechanism to deliver content, functionality, and excitement to users and customers (and, no, browser-based apps won't rule the world - sorry).

Imagine Sun without these client-side technologies and products. A company focused solely on providing web/cloud server-room infrastructure - I'm sure that's a viable business in the short-term - but will it allow Sun to run a differentiated, profitable, growth-oriented technology business in the long run? Or would Sun simply become a check-mark item for commoditized back-room technology? Best-of-class maybe, but still commoditized - a white-box Google or Amazon for web infrastructure? And then what?

Pole Position

I actually happen to think Sun is in a very good position - a portfolio of innovative and must-have technologies and rapid adoption through open source has created a host of opportunities. What Sun needs to get better at is to be selective, focus on the opportunities, put appropriate programs and products in place that make it easy to do business with Sun, and then execute aggressively to provide the market what it wants.

Killing the very client-side technologies that have played a big role in positioning Sun for these opportunities in the first place seems short-sighted, at best.


-- Terrence


Without Java, SUN would be an hardware company among many other ones. Without hardware, SUN would not have enough profit centers, not enough opportunities. Both parts can't be separated, as if one has problem, the other can help for profit. I think the same is true for both sides, the client-side and the server-side. Maybe the business model should be better emphasized. When I click on the link "Price and Buy" for Sun GlassFish Enterprise Server (, I go the page that says : "" Get Sun GlassFish Enterprise Server Download the Sun Java System Application Server at no cost no kidding. "" Then, the business model does not appear easily. I think both sides, open source and business model on top of it, should be equally emphasized, and may be better presented on your web site. While I think following open source is the way to go, no doubt here, I don't think being called an "open source company" helps that much, if people believe everything should be free. Why not putting also the accent on "no lock-in" or "no lock-in from standard compliance" ? I know the SUN softwares, but I don't know the corresponding business models. Is licence sold by CPU ? Is subcription sold by SLA ? by a number of bug support during one year ? etc. Years ago, SUN had no reputation for its softwares. Now, Glassfish, for example, is known of being one of the best, if not, the best Java EE servers out there, open source or not. I think you have a great portfolio and reaches a critical threshold that could bring opportunities, while mixing thoses technologies or while having enough big communities. Anything that could push bigger those communities (pluggable/extendable architecture, docs...) will be also a plus here.

Since Rich mentioned Apple, I'll bring up the oft-repeated speculation about a Sun - Apple merger. It would make good sense for Sun, but why would Apple do it? To become more than a consumer products company, which is definitely a commodity business domain. Sun would benefit from Apple's marketing savvy and ability to identify opportunities (iPod, iMac, iPhone). Apple would diversify into Sun's back-end businesses. The two complement each other with little overlap. Apple would provide the client side and Sun the server side. So how do two businesses that are in different commodity business lines make money together? By selling the client side and server side together along with high value-added professional services. The problem is that neither company has been very good at selling services. Sun surely sees the need to change, and Apple should recognize the opportunity to diversify. But it probably won't happen, at least not with the current leadership at both companies.

Hi Terrence, I think the trick is not fighting commoditization but rather, surfing that wave, as its going to come in and you will either ride it to the beach, or drown. But if you are going to do that you've got to keep swimming back out and catching waves - inventing really new stuff, making high margin profits early in its life cycle, then making low margin profits on large volumes late in its life cycle. Sun rode the wave twice - once on commercializing advanced microprocessor based workstations and servers, and once on Java. But the Java wave started in 1995 and has pretty much beached now, and Sun did not catch the next wave. On client-side - I think you're right, that there's opportunity there. Apple has certainly found some. The trick I think though isn't inventing the tech, it is monetizing it. What should Sun do to monetize JavaFX, for instance? You can find my ideas on What Sun Should Do at