July 2007

There seems to be a lot of discussion about whether Project Indiana at openSolairs is a Linux clone or not.

The Project Indiana website states:

“Project Indiana is a new project to create an OpenSolaris binary distribution. This distribution will focus on providing a single CD install with the basic core operating system and desktop environment, with the opportunity of installing additional software off network repositories.”

Ian Murdock, of Debian fame, now with Sun, says:

“It is not a Linux copy thing,” …. “It’s a best of both worlds thing.”

Most folks responding to the story at Slashdot and apparently those in attendance at an OpenSolaris user group in NYC (which fellow IBMer Mike Dolan attended) think Project Indiana is either (1) a Linux copy thing that should Sun should give up on, (2) not something that Solaris customers want, so Sun should stop wasting time and resources on it, or (3) both 1 & 2.

Jack Loftus writes:

“Now, from what I can understand the problem here is that Sun has continued to competitively position itself with Linux users with features like DTrace and ZFS when perhaps they should take a page from the Linux playbook and start siphoning off some Windows market share.”

Hint, it’s because the guys that buy/run/administer Linux servers are typically not the same guys that buy/run/administer Windows servers. Linux hurts Solaris (& other Unix OSes) much more than it hurts Windows. Sun has realized this and is trying to win over the “low hanging fruit” who used to be, or could be, Solaris customers.

What’s wrong with Sun trying to build a Linux clone with OpenSolaris? Why shouldn’t Sun compete with Linux if that is what Sun’s strategy calls for? We participate in a free market, let Sun compete as they wish. The strength of the Linux ecosystem, backed by a truly open and extensive community should take care of the rest.

Nothing personal here, it’s just business.

PS: I should state: “The postings on this site are my own and don’t necessarily represent IBM’s positions, strategies or opinions.”

Earlier this week I wrote a “what if” post suggesting that Microsoft buy Red Hat. The comments here and on Open Sources at InfoWorld were lively to say the least.

Many comments were a variation of the “wow, you’re clueless” theme. I approved every single one of these comments.

Some readers suggested that it would never happen because of the DoJ & EU. A valid point that I hadn’t even considered when I ended the post with:

“PS: I truly doubt this deal will ever happen, but it’s interesting to think about the possibilities.”

Private email discussions I had on the topic suggested that Microsoft’s culture would severely limit any hope of a positive outcome. Valid point. But corporate cultures evolve.

I was somewhat surprised by how strongly readers debated the very notion of such a deal based solely on Microsoft’s history towards open source. Don’t get me wrong, Microsoft has done plenty of clueless things in various open source areas. But, to judge the possible outcomes of a Red Hat acquisition based only on history is, I humbly suggest, missing the point. (Again, keep in mind that this acquisition will likely never happen, but stay with me for a second longer). For Microsoft to even consider this acquisition, they would have to dramatically change their historical views on open source. For companies like Microsoft, change is a lot easier than holding steadfast to outdated strategies.

Change doesn’t happen overnight. However, the impacts of a threat like open source are not felt overnight either. It’s taken Red Hat nearly 15 years to get to where they are today, and the open source movement has over 40 years under its belt. The results have not been industry shattering on either front. I am not minimizing the success of Red Hat or the open source movement by any means; I’m only putting them into perspective.

Sometimes we build deep rooted views about technologies and companies that don’t allow ourselves to consider situations in which the technologies or companies evolve. Is it really “clueless” to think that over the next 5 to 10 years Microsoft could increasingly adopt open source strategies into their broader corporate strategy?

While buying Red Hat may never happen, what if Microsoft launched a Linux distribution, maybe even based on RHEL? Ahh, but what about Microsoft patents and GPLv3. True, but what if Microsoft realized that the revenue potential from shipping “Microsoft Linux” is 2x or 10x the revenue potential from licensing their “Linux patents”? These are the types of decisions that companies like Microsoft make daily. The point isn’t whether Microsoft, its culture, or employees hate Linux/OSS. The point is that Microsoft, its culture and its employees are interested in the future success of the company and making as much money as they can. To date, Linux/OSS has been seen as a risk towards this goal for Microsoft. In the future, Linux/OSS may well be a driver towards this goal.

Change happens.

PS: The *most* surprising thing about my post is that Dave Rosenberg had a similar post encouraging Microsoft to buy JBoss and/or Novell….back in 2005. Sigh…and I thought I was an original thinker…. :-)

PPS: I should state: “The postings on this site are my own and don’t necessarily represent IBM’s positions, strategies or opinions.”

Some background:

  1. Matt asked “Why doesn’t Oracle just buy Red Hat?”
  2. I explained why Oracle would not buy Red Hat
  3. Luis Villa replied to Matt’s question: “Because Red Hat employees would leave en masse.”
  4. Microsoft announced fiscal 4Q07 growth of 13% on Thursday (or 16% if you only count their true software revenue – which falls into the “Client”, “Server & Tools” and “Microsoft Business Division” reporting categories). Microsoft crossed the $50 billion total year revenue mark with the close of fiscal 2007.
  5. I compared Red Hat’s stock performance over the past year vs. some Traditional software vendors (see below)

If you’re still with me….

Red Hat is growing and executing well. Financial analysts expect Red Hat to hit $517M this year (fiscal 2008, ending Feb. 2008), and $631M in fiscal 2009. At this pace, Red Hat should cross the $1 billion revenue mark in fiscal 2011. Red Hat may well be the gorilla in the Open Source marketplace. But after everything is said and done, that marketplace is tiny in comparison to the total software market. If you believe in the stock market’s ability to predict a company’s future value, one could argue that Red Hat investors are in a “sit tight” mode right now. At a P/E of 72 and PEG of 1.44 (vs. Google’s PE of 45 and PEG of 0.99), Red Hat’s stock has likely priced in as much growth and “great news” that we could think of. Few doubt Red Hat’s position in the overall OSS market, but some may be waking up and asking whether being #1 in 1.8% of the software market is enough to drive the multiples that Red Hat shares enjoy today.

While both vendors have strong operating system franchises, Red Hat isn’t really eating into Microsoft’s revenues. IDC predicts that the Linux & Windows markets are growing 26% and 9%, with Unix revenues declining by 3% from 2006-2011. Red Hat’s Jboss division adds a JEE portfolio that does compete with .NET as the infrastructure for enterprise applications. But here again, it’s very unlikely that Microsoft faces off against JBoss in (m)any customer deals. It’s more likely that Microsoft competes against IBM WebSphere, BEA WebLogic or Oracle AS, and JBoss only comes into the picture when the customer has already selected JEE. While there is some overlap, Red Hat is much more complimentary to Microsoft’s offerings than we’d like to think.

Just imagine a Microsoft that could offer customers a choice of Windows/.NET, Linux/JEE or, and here’s the magic, BOTH. The fact is most customers have heterogeneous environments, and those that don’t today, will likely in the future.

In the face of OSS competition, one of the best moves we made in the IBM WebSphere division was purchasing Gluecode. As I’ve mentioned (over and over), having a free application server, based on the open source Apache Geronimo project has done nothing but spur the growth of our overall WebSphere Application Server family. In some cases, the customer chooses WAS CE, in other cases, they choose Traditional WAS products. We help customers be successful with their choice and, equally important, ensure that their previous investments in WebSphere infrastructure are protected. This is exactly the scenario that Microsoft could create for themselves. Microsoft would be able to offer Windows, Linux, .NET or JEE in various combinations to solve customer problems. As a competitor, this would be a scary combination.

What’s more, Red Hat could help Microsoft gain OSS street cred almost instantly. To me, this would easily become Red Hat’s most important contribution to the software industry. Forget being #1 in 1.8% of the software market. How about helping a $50B software company evolve its thinking around OSS in order to become a $75B software company while increasing customer choice and satisfying customer needs? Goosebumps.

I doubt this acquisition will ever take place for three reasons. First, because of vendors such as IBM, HP, Intel and Oracle who have investments in/with Red Hat. Second, because Microsoft wouldn’t want to take the risk. I’d argue that there is much less risk than appears on the surface. Sure, there would be some internal friction during product positioning discussions if the deal went through. But internal friction is healthy and shouldn’t get in the way of helping customers succeed with your offerings. Lastly, as Luis’ comment highlights, Red Hat’s culture would also pose a barrier to this deal. But I’d argue that the only thing that these comments do is put a damper on the deal price, which, at the end of the day, is bad for Red Hat investors. One could argue that JBoss employees felt the same way about Red Hat prior to the acquisition. And yes, some left Red Hat, but some stayed. It would be up to Red Hat management to convince employees about the historic importance of their efforts inside of Microsoft, which would be a pretty compelling reason to stay.

A Red Hat marketing slogan states: “truth happens”, what about “change happens”? And why not help Microsoft change?

PS: I truly doubt this deal will ever happen, but it’s interesting to think about the possibilities.

Zack asked:

So perhaps a question that might be worth asking is if Windows and Office only costs $3 in China, how sustainable are it’s prices in other markets?

I remember reading about multinational companies consternating about selling online in the early days of the Interweb. Most companies sell at different prices in different markets. But until the dawn of widespread e-commerce, purchasers in different geographies couldn’t easily compare prices across geographies. Today, I can easily find the price of an iPod for American customers vs. here in Canada, but I’ve never checked. I know that the US price won’t change my decision or acquisition strategy.

Is a $3 Windows & Office package for students going to have a marked impact on comparable prices in other parts of the world? Sure, to some degree. But I’d argue very little. The growing concern about open document standards may likely have a larger impact.

Now to the larger question of Microsoft’s strategy in India (& China):

This is based on a sample size of 2 cousins in India. One completed a business & IT degree and the other completed a degree in computer engineering. As they tell it, having a MCSE designation is critical to getting many IT jobs in India.

My cousins have grown up with Microsoft Windows & Office in their homes and in the local internet cafes. Most kids learned to use Office & Windows in university. Because of their educational backgrounds, my cousins also learned Visual Studio, VB, SQL Server and fun stuff like administering W2K.

Playing Microsoft Strategist, I’m happy to sell Office & Windows to students for $3 if it means I have them hooked on an end-to-end Microsoft technology stack. Training the next generation of workers on Windows & Office makes it more likely that employers will purchase Windows & Office. Don’t underestimate training & support costs that can be minimized.

No here’s the kicker…the kids love Microsoft. We may want to believe that OSS would be best aligned with emerging countries especially because of the code freedom and lower costs. When I questioned my cousin why they weren’t using Linux at his company (Large bank), he asked: “Is Linux as secure than Windows?”

Again, remember this post is based on a sample size of 2.

Some of you know that I’ve decided to track the demise ;-) of the traditional software market on a quarterly basis using IBM’s WebSphere branded revenue as the basis. I use the WebSphere division for no other reason than it’s the part of IBM that I report into. Here are the 1Q07, 4Q06 and 3Q06 posts if you fancy. IBM announced 2Q07 results today.

A few points of interest from 2Q07:

  1. IBM Software grew at 13% (or 9% at constant currency: i.e. if currency exchange rates were fixed to equal their 2Q06 rates)
  2. WebSphere branded middleware grew at a very healthy 28%
  3. IBM’s other Software families grew 12%, 11%, 33% and 21% for Lotus, Rational, Tivoli and Information Management respectively

I stress very healthy because 28% growth is pretty awesome when you’re growing from such a large revenue base, which unfortunately, IBM does not make public. But, if you have access to Gartner or IDC data, you can easily get to a ballpark number of total WebSphere branded revenue.

WebSphere Branded Middleware Quarterly Revenue Growth:

Quarter Y/Y Qtr Growth From:
1Q04 24% Source
2Q04 N/A Source
3Q04 14% Source
4Q04 18% Source
1Q05 11% Source
2Q05 18% Source
3Q05 14% Source
4Q05 4% Source
1Q06 26% Source
2Q06 17% Source
3Q06 30% Source
4Q06 22% Source
1Q07 14% Source
2Q07 28% Source

IBM WebSphere Application Server Results:
Some might reply: “come on Savio, you expect me to believe that in the face of open source competition, the application server business is growing at all? Maybe WebSphere Branded Middleware is growing, but the underlying application servers surely aren’t, right?”

IBM provided WebSphere Application Server’s yearly growth in 2006, so we can compare that figure to the growth of WebSphere branded middleware in the table above:

From pg. 34 of IBM’s 2006 Annual Report :

“Revenue from the WebSphere family of products increased 23.3 percent (22 percent adjusted for currency) and was led by doubledigit growth in WebSphere Application Servers (25.3 percent) and WebSphere Business Integration (22.7 percent) software versus 2005″

Translation: The WebSphere Application Server family revenue growth outpaced the figures that you’ll find in the table above.

Yes, even in the face of open source competition, the WebSphere Application Server business is growing at 2-3x the market. How? We’re helping our customers address their business needs to achieve success. In some customer situations, we help customers achieve success through the use of WAS Community Edition (WAS CE). In other situation we help customers achieve success through the use of the traditional WebSphere Application Server family. The results are pretty clear.

Just a heads up that Red Hat’s Bill Burke (of JBoss fame previously) is blogging here. Bill does write about techie stuff at the JBoss Matrix, but this blog seems to go beyond the bits and bytes and get into things like:

Should you start an OSS project at Apache?
Should Interface21 join the standards process?

I’ve had mild disagreements with Bill in the past, but he’s always made me think….and his post on “Apache Business Model failure” is no exception.

I had never thought about “who owns the brand for an Apache project”. Now that Bill points it out, the answer is obvious.

It’s clear that established software vendors prefer neutral communities such as Apache or Eclipse. These vendors seek to jointly develop some widget/product/project and then use it within a commercial offering under their respective brand. As a result, “who owns the brand for an Apache project” is not a very salient question.

I’ve argued in the past that users benefit from projects developed at Apache (or Eclipse) as multiple (competing) vendors and parties have a stake in the direction of the project.

Maybe we’re going to see Apache become the home for OSS projects that are sponsored, in one fashion or another, by Traditional vendors. Additionally, Apache could remain the home for implementations of core technology by individual developers who have a vested interest, but aren’t doing so as their primary source of income.

Note however, I don’t think Bill’s comments have any impact on the importance of the Apache Software License.


Matt asks (almost as an aside to a post on a related topic):

My question: why not just buy Red Hat? Before Red Hat buys MySQL, and gives those database numbers a run for their money?

Then, Larry Dingnan writes:

A Credit Suisse analyst thinks BEA Systems is likely to go on the auction block in the next three to six months.

And the potential buyer is two likely suspects: Private equity firms or Oracle, which indicated it has no plans to slow down its acquisition pace.

I’ll play Nostradamus and predict that Oracle will not buy Red Hat. Oracle may buy BEA. Here’s why…

If it’s possible to summarize the complexities of decisions in a large business such as Oracle based on reading a book about its founder & Chairman, then Oracle won’t buy Red Hat for the same reason that Larry didn’t buy Netscape. And no, it has nothing to with whether Larry’s cat can write an operating system. It does have everything to do with the “value” that Red Hat brings to the table. Essentially, what does Red Hat have that Oracle doesn’t?

Three key reasons for buying a tech company would be: technology, brand recognition and customers.

By announcing Oracle Unbreakable Linux, Oracle has already proven that Red Hat doesn’t have a whole lot of technology that can’t be easily replicated. Sure, Oracle UL hasn’t been a rip roaring success to date, but I agree with Matthew Aslett, that this is a long-term investment.

Red Hat does have an awesome brand. But its brand is tied to open source. As long as customers relate the two, and the OSS market remains less than 1.8% of the total software market, the value of Red Hat’s brand to Oracle is questionable.

Additionally, Red Hat doesn’t bring a whole lot of customers that aren’t already Oracle customers in some shape of fashion. Even if Oracle were to migrate every RHEL customer running Oracle DB to Oracle UL, the end game isn’t getting customers to Oracle UL. Oracle’s goal is getting customers to Oracle middleware & Oracle apps. Customers don’t typically make application & middleware stack decisions based on their operating system. Consider the Microsoft customers who run WebSphere or SAP or Oracle Apps on Windows.

Oh, and if Oracle tried to buy Red Hat, I’m sure that other vendors interested in the future of open source and Linux would have some role to play in the discussions.

BEA is different. They have IP that Oracle can’t get their hands on without an acquisition. BEA has good brand recognition in the middleware market, an area that Oracle wants to grow. BEA’s customers are likely also Oracle’s customers (as BEA always had an affinity with Oracle DB vs. DB2). But BEA’s middleware customers could be migrated to Oracle’s middleware stack, which is strategic to Oracle.

PS: My admiration for Oracle will subside after I finish the book I’m sure :-)

Next Page »