Do you plan to expand the use of Linux at Hess?
Its not like were rolling out Linux across the company. The Linux cluster is only in two locations now: Houston and London, and may go into Southeast Asia in the future. We take good ideas and we look for places where theyre applicable. The sort of problem that Linux clusters can solve is very idiosyncratic. Its parallel processing, which involves a trade-off between a supercomputer and PCs. Were not moving to Linux generally in the company because theres no inherent benefit. We have a very standard Microsoft environment at the desktop and the server. What we have now is a sunk cost in Microsoft. Our analysis has indicated the total cost of ownership of the Linux environment for general business applications is greater than Microsoft. You need people trained on Linux, you need system management tools, you are creating an inherently more complex environment with multiple platforms. We take great pleasure in reminding Microsoft that there is an alternative, but until the economics get much better, we wouldnt move there.
How do you see the Linux movement evolving?
Over the next few years, I believe the open-source movement will be more about Linux versus Unix than anything else. To that end, at Hess, Linux will assume a larger and larger percentage of our total Unix application portfolio as we switch out of the proprietary Unix systems. This trend will be modified somewhat by the fact that we are seeing at least one of our major engineering application vendors porting its stuff to Wintel, so I expect a sizeable chunk of what is now on Sun Solaris will eventually move to Wintel.
Longer term? Its harder to predict. If nothing else happened, I would expect Linux to remain a niche operating system. However, the decision by the Chinese and several European governments to force the issue and migrate their office technologies to Linux will add weight to the product that it would not otherwise have had, and likely produce a more robust general business suite. If this happened, we would look at the value of those products over whatever we were currently using. But this is certainly not assured.
Unix had a similar backing and never went anywhere relative to Microsoft. This is the Achilles heel of open source: When anyone is free to change it, you get a lot of versions and hence fragmentation. Operating systems are infrastructure and generally not a competitive advantage. You need them to work with everything and to be simple. Unix had the hope of that but lost it when the hardware manufacturers started adding proprietary extensions. This is already happening with Linux and will be a limiting factor in its market uptake.
The real advantage of our Linux cluster is that it is a cheap parallel-processing environment and not that it is Linux, although the cost would certainly go up if we had to pay licensing fees. Migrating our entire business infrastructure to Linux would require a huge investment in tools and skills, and at best we would have what we have today: infrastructure that works.
At this time there is at least an even chance that we wouldnt get that reliability and the degree of automation that we have today. Why in the world would I want to do that with something that has no competitive advantage? Sure, we all hate the thought of single vendor lock-in, but lets face it, we benefit from it. We did an analysis recently that showed, on a constant dollar basis, that the Microsoft suite of products has not increased in price for us over the past five years. If you add into that all the functionality that has been added, I think we are ahead of the game. On the other hand, Microsoft certainly isnt the only game in town for the latest wave of functionality, such as Web services, content management, wireless e-business and so forth. We have done work in these areas and gotten very good results without Microsoft products. Where we go in the future will be a function of how competitive they are.
Is Microsoft trying harder now, because of Linux?
Microsoft has worked hard to craft a value-added relationship with us. A few years ago they would show up when it was time to collect more money from us. Now they are all over us, offering migration help and opening up their network of accounts to help us explore new applications. Its up to us how we value these things, but it is certainly a positive move on their part.
Richard Ross Career Highlights
Richard Ross Career Highlights
Amerada Hess Corp.
• Vice President, CIO
McKinsey & Co.
• Consultant (London)
• Director of IS
• Project Manager (Tokyo)
• Internal Consultant, Office of the Chairman
• Manager, Telecommunications Projects
• Program Director, Industry Services
• Development and Manufacturing Engineer
• Manufacturing Engineer
• MSEE/Computer Science, Northeastern University
• BSEE/Biomedical Engineering, Washington University