Road thoughts on Microsoft, Oracle, and Open Source

Many of you can appreciate the weariness of traveling, waking up in a hotel room and having to think for a moment where you are. Or sitting in an airport lounge forgetting what state you are in. Such is the condition I find myself in as I write this. But, I’ve been asked so much over the past few days what I think about all this Microsoft and Oracle stuff and its affect on open source. So, for the record, here are my thoughts.

Giant corporations like Microsoft and Oracle have a lot going for them. For one, they have many great employees with a ton of talent. Secondly, they have a lot of marketing pull. The stones (boulders) they drop into a lake make much bigger ripples than the tiny pebbles from most other companies. However, with such greatness of size and power also comes a certain amount of antipathy from the masses. Notwithstanding Wall Street’s view, it’s almost as if there is some fundamental principle of external disenchantment that sets in once a company passes a certain threshold. Regardless of one’s world view, it is a fact that both Microsoft and Oracle recently dropped some pretty heavy stones into the open source lake.

So, what does this all mean for the open source community? Not only do I think it is good news, but I also think inevitable. Let’s start with the inevitability and then move on to the goodness.

Oracle wants to be in the operating system environment, actually the whole stack environment (think Fusion) for their business process architecture. They had been rumoring at Linux for a long time, so this announcement should have been no surprise. Oracle wants to be “the” platform for business processing, and the operating system is a natural part of that architecture. Now, how much Oracle will really do with Linux remains to be seen. They certainly won’t want to fork RedHat unless there is a really solid business justification behind it. Some may speculate a “conspiracy theory” to depress RedHat’s market cap, but while highly juicy from an editorial perspective, I highly doubt that is what drove Oracle to make this move. The OS software is free, and Oracle’s focus is to do everything they can to drive businesses to their application environment. (NB: I will say that it is rather interesting that you don’t find much recent information from the Oracle execs on their blogs. A quick look into a few of them revealed last updates from over six months ago.)

Likewise, Microsoft is in the operating system environment. And, what many people probably don’t know is that Microsoft has been working with Linux for quite a while. Bill Hilf, a really smart guy who came from IBM’s Linux team, runs the Microsoft open source team, and these guys have been quite busy. And, of course, Microsoft wants clients to use their software, so why would they ignore a huge market in the Linux space. Now, how many Novell “coupons” Microsoft actually ships also remains to be seen. But, Microsoft will certainly continue to find the best balance between the revenue they generate from their operating system and the opportunities (both from a revenue and PR perspective) with Linux and the open source community.

So, the reason I think both of these “moves” are positive from an open source perspective is because they add additional legitimacy to the whole space. Of course, all of us already swimming in the open source lake know that the water is nice and warm, clean, safe, healthy, and here to stay. But, for those looking in from the outside at this (dare I say) oasis, the obvious question is, “Is that real?” When companies like Microsoft and Oracle start splashing around, you know you’ve crossed a tipping point. Now it no longer looks like an oasis, or even a lake. It is starting to look like an ocean.

West Coast travels and SOA

It’s a long trip travelling east-to-west, fighting the jet stream along the way, from PHL to SFO. However, it affords the traveler a great chance to catch up on a ton of reading. Since I travel quite a bit, I am the kind of person that keeps a stack of material to read for just such an occasion. This trip allowed me to get through a two-inch thick manila folder of great open source articles (and about 10 packs of US Airways’ “Fancy Cashews & Honey Sesame Sticks”).

The topic I’d like to throw open for discussion here is the correlation between open source and services-oriented architecture (SOA). I’ve come across a few analysts that don’t see much (if any) correlation between the two, but thankfully they seem to be in the minority (and I won’t mention them by name). Open source is an enabler to SOA, and although not all SOAs are built completely with open source components, the modular, loosely coupled nature of open source components make them ideal for many SOA implementations. And, with approximately 80% of CIOs making SOA a priority, this is a worthy discussion.

But, a step back is probably in order. If you ask 10 people to define SOA, or even “architecture” for that matter, you will most likely get 10 different definitions. My view (or definition) of SOA in one sentence: a defined, governed infrastructure environment (or parts thereof) where business rules & processes have been decomposed into a set of registered, discoverable services, which, when invoked, perform a function and perhaps return a result. The goal: leverage existing assets (ie legacy applications and mainframes) and the build or acquisition of new assets in an intelligent, modular approach such that component reuse is high and solution developers can focus on their core value-add rather than having to reinvent parts of the wheel for every solution. Analyst firms like Kennedy estimate that organizations can save 30-40% on their infrastructure costs “because clients can easily reuse software once it is developed”. But another huge benefit of building a SOA environment and thinking through the business processes that are core to the business is the ability to more effectively link business rules with the infrastructure and thereby create a paradigm where the business rules drive the IT infrastructure, rather than the other way around. At Unisys, we call this 3D-VE or 3D visible enterprise, where the linkages are defined all the way from a company’s vision, down through its business processes and applications, and into the IT infrastructure. By seeing how it all fits together (or, in many cases, doesn’t) and understanding the impact of changes before they are actually implemented, it is a lot easier to build a future state environment like the one I just described.

The hard part, of course, is developing a governance model that is used by the entire organization and coming up with the right level of “granularization” for the service components. Take the presentation layer as one example. User sign-on. Security. Policies and access models. Interoperation of the security components with the rest of the middleware environment. Companies need to think through all this to get it ideal, and it will have ripple effects throughout the infrastructure. But, one of the beauties of open source is that you can start with a low cost without spending a fortune and being stuck on the path of a purely proprietary implementation. The hottest SOA areas right now are around portals, integration, business intelligence, and security, and there are plenty of open source components to help.

Sun open sources Java

Kudos to Jonathon Schwarz, Rich Green, Bill Joy, and the rest of the Sun team for more than just following up on a promise but making a very difficult business decision to release some of their crown jewels. I suspect that the resultant code base will be enhanced with new features and that more developers will be attracted to Java. It should be fun to watch.

Open Source, The 3 C’s, and a New Beginning

It starts with an idea … as always. Although I’ve commented extensively on other’s blogs and been quoted quite a bit in the press for what I’ve been doing around Open Source, it’s time to put together a new model for such “communication”. One in which my perspectives and insights can be shared, discussed, and debated directly without the other “noise”. But, perhaps even more importantly, a communication model in which I can get honest feedback from the global community and help drive our shared passions forward.

Why do this? In the words of the great essayist Emerson, “If we are related, we shall meet.”

I am passionate about a lot of things, and one of them is Open Source. Not just because it’s such a disruptive technology and is turning the software world upside down, although I do think that is cool. Not just because it affords businesses tremendous cost savings and flexibility improvements, although that’s precisely what our solutions at Unisys do. But because the open source development model fosters what I call the 3 C’s: community, collaboration, and components. People from all around the world, coming together to design, build, test, release, and support standards-based modules in an open-forum environment. But, for me, it’s much more than “just software” (or “free software”, although that’s a topic for another posting) … it’s a development model that is revolutionizing the way people build solutions to problems.

By leveraging like-minded souls across the globe who share a common interest and passion, solutions can be built much more quickly and efficiently than “typical” development projects. A perfect example comes from the Goldcorp gold mining corporation that needed a way to improve their success at locating gold in their mines. By sharing their geological data on the internet and hosting a contest to offer a large sum of money to the team that could process that data and find the right spots to mine, Goldcorp became hugely successful. Mind you, a mining company’s geological data is the key to the kingdom. Imagine the fortitude it took for Goldcorp’s CEO, Rob McEwen, to take that bold step. Rob credits his idea to the Linux development model he had recently learned about from a seminar at MIT. Sure enough, people from around the world competed, and some teams developed new methods for processing the geological information (including new methods for drilling) that Goldcorp’s team had never even imagined. An Australian team from half-way around the world from Goldcorp’s headquarters in Ontario won the big prize, and they never even saw the gold mine. Check out the article Don Tapscott wrote on Goldcorp last year. As an aside, I met Don Tapscott at Gartner’s Open Source Summit a month ago in Phoenix. His visionary insight into technology trends and evolution are superb, which he’s translated into many books worth reading.

So, as the person who started and runs the Open Source business in Unisys, I can assure you that the development model I just described works equally well in giant corporations like ours (think 33,000+ employees, think over 100 countries around the world, think over $5.7B in revenues … starts to give a sense as to the magnitude). But certainly not without challenges. There is a lot of education required to create such an environment, and a really good governance model in place to make it work well. However, with the right focus and leadership, it can pay off immensely (as it has for us and others like Goldcorp).

In essence, open source is successful and will continue to revolutionize the industry because it leverages the collective genius of the community and helps them collaborate more effectively thus resulting in globally reusable, high-quality components that they can continue to build upon as a community. That development model is here to stay and will expand into new territories. It’s great fun to be a part of this journey, and thanks for joining me on the ride.

So, back to Emerson, if we are related and it’s time for us to meet, I can be reached at anthony dot gold at unisys dot com or privately at apgold at yahoo dot com. I promise to respond to every email message as well as read every comment in this blog. I look forward to making your acquaintance.