Q&A: Red Hat CEO sees user focus on architecture, stability

Computerworld - Open-source software will be a hot topic in the month of February, between next week's OSDL Enterprise Linux Summit in Burlingame, Calif., and the LinuxWorld Conference & Expo, which begins Feb. 14 in Boston. With that open-source focus as the backdrop, Red Hat Inc. CEO Matthew Szulik spoke recently with Computerworld about what's most important to his customers, the Raleigh, N.C.-based company's upcoming 4.0 release and the competition.
Excerpts from that interview follow:
What user trends have you been noting? What oftentimes gets missed is that customers are increasingly deciding on an architecture and not on a point product. The leading companies that we're doing business with are all focused on how do they get to a standardized, commodity Intel architecture as fast as possible.
Most of them are migrating from something, whether it be an IBM mainframe or one of the 61 Unix variants. Outside of the United States, it's increasingly becoming Microsoft platforms that are finding their way to an industry-standard, Linux/Intel combination.
What we've witnessed increasingly in the last two years is that the people who brought in Linux and open-source software are now starting to move into more mission-critical deployments, and I think that's not an accident. ... Of course, [Red Hat Enterprise Linux] 3, the product that's currently in the marketplace, was a very good release, and that was really a major platform to get ISV [independent software vendor] support. All of your core large enterprise applications from Oracle, IBM, BEA, Computer Associates, etc. run on RHEL 3.
That combination is giving enterprises assurance that this migration is relatively safe, and there is a very limited downside. This will be an exciting year because I think you're going to start to see very large customers talking about some quantum return on investment.
What's been happening on the client side? The next move, I think, we'll start to see is the change out of the desktop. Many people have a conception that it'll operate in a similar paradigm. They'll have a hard drive, and why not swap out Microsoft Office for another Office? But the last four years, we haven't believed that because we've been building out our own desktop implementation focused on how to solve the security problem, the manageability problem, knowing that the thin-client implementation [is] able to provide functionality like single sign-on, better authentication, better management. [Those] were really the ways that our enterprise customers wanted to approach the desktop -- not, "Is somebody's spell checker working better?"
What are customers asking for from an architectural perspective? A complete software stack? A development architecture?