HP exec on navigating the open-source waters
- 15 May, 2006 09:27
Christine Martino was appointed vice president of Open Source and Linux Organization (OSLO) at Hewlett-Packard Co. (HP) in November, succeeding Martin Fink, who took over HP's non-stop server group and remains Martino's boss. An 18-year veteran of the company, Martino often runs the widely-dispersed group out of her home office in California. when she's not traveling. She talked this week with Computerworld's Eric Lai about inking deals for HP to sell and support popular open-source applications such as OpenLDAP, JBoss and MySQL in the last 18 months, but said she is now less focused on alliances and more on making the software work better for customers.
Your organization is part of HP's Enterprise, Storage and Servers division, yet it works in all areas of HP where there is open-source or Linux activity. How do you measure what contributions you make to HP's fiscal bottom line? And what was that contribution?
That's not an easy thing to calculate. We look at a number of different factors -- our Linux server sales, for instance. We have enjoyed No. 1 market share since it's been counted. We also measure how much in Linux and open-source services we help drive. There are some softer measures [such as] how many deals have we helped because we have open-source knowledge? Because of our team, HP is a real contributor to the open-source community, as opposed to being just a consumer. That has a sales benefit. Enterprise customers want to know that the vendor they are entrusting their business to knows how to work with the open-source community.
One analyst compared Hewlett-Packard to Switzerland, because you can play in the Unix, Linux and Windows areas simultaneously without offending any of your partners, such as Microsoft. Meanwhile, your services group cleans up.
I like being called the Switzerland of the high-tech industry. It applies in operating systems but also up the stack, as well. Our story is one of choice. Our goal is to have a level playing field across platforms, from HP-UX to Windows to Linux. I know very, very few customers that have a homogenous environment. They have legacy systems to deal with, new applications going up. You can pick the OS that works for you. It's truly open choice, because we don't have a proprietary software stack that we have an ulterior motive to try and steer you to.
IBM is the big company most identified with open-source, more than HP. Sun has made Solaris and other pieces of its technology open-source. And the new CEO Jonathan Schwartz with his ponytail looks like an open-source guy. Is there any danger Sun will eclipse HP in open-source?
I'm not very worried about Sun eclipsing us in the mindshare battle or any other one. Our strategy works for customers because it offers them what they need.
Will you be selling and supporting any more open-source applications in the near future?
For the initial middleware stacks we're focusing on, we're in pretty good shape. We wanted to have a directory services stack, a database stack and a J2EE stack. We've got the pieces together for that now, and that's what our customers have wanted so far. We'll probably add more components, but right now we have a pretty good mix. What I'm telling the team to focus on is tying these stacks of open-source middleware components to specific commercial application areas or things higher up the stack, like SAP, for example, or specific vertical applications such as key financial services applications.
Basically, we have our dream-team lineup now. We'll add more applications as customers request them. But tying the applications to specific areas is the top priority now.
If companies like Oracle and IBM keep snapping up open-source companies, aren't you at all worried that you won't have anyone left to partner with?
I'm not too worried. Take Red Hat buying JBoss. We were already working with both of them so that had no effect. Oracle is a very strong partner of ours, so that doesn't break anything there. But fundamentally, I don't think it's possible for a large software company to go out and buy up all the open-source companies. The community will fill the hole because the community development model is proven and here to stay.
Would HP ever get in the game and start acquiring open-source firms?
It's not a high priority for me right now to buy open-source content across different products, because our strategy of choice is from OS layer to middleware. That has served the company very well. I don't see the benefit of breaking that strategy in either the open-source or commercial software side.
There are still many gaps in terms of software - even drivers - available for Linux users. How does HP help fill that gap?
We do a lot of work with Kernel.org and the Debian community. We try to fill those gaps all of the time. I tell the R&D team, 'Anything that you can do on HP-UX, we should be able to do in Linux. That's your goal.'
HP sells PCs installed with Mandriva Linux in Latin America. When could we start seeing such Linux desktops elsewhere?
We could do that anytime. We do certify Novell on our desktops. We don't ship desktops with Linux now because the market is not big enough or cohesive enough from a worldwide perspective. We watch it closely, though, as does the Personal Systems Group [PSG]. Our regional groups have the freedom to work more locally if a customer wants it. So we do Mandriva in France and Brazil, and Ubuntu in South Africa. If there is big interest when SUSE Linux Enterprise Desktop 10 from Novell is released, the PSG can turn on a dime and offer it pretty quickly worldwide.
I've heard, though, that OEMs are reluctant to install Linux because they get such great marketing dollars and softer incentives such as support from Microsoft when they install Windows. Also, Linux doesn't have a track record of adding new features and making the PC hardware obsolete every three years like Windows. So in the long run, supporting Linux is not smart for the PC vendors. I'm not sure I would agree with that. Certainly, Microsoft is big player in this space and a big partner of ours for desktops and there are incentives that come with that. But we made the investment to put Linux on a level playing field in the server space. I would assume that when the global market is more there for us to jump more deeply into Linux on the desktop, we would do it -- making that same investment, so it would become a real choice for our customers.
But isn't the desktop market much different than the server market?
It requires volume and churn and has really low margins. That's true, I see your point. But at some point perhaps -- and I'm not predicting this, because I don't have a crystal ball -- if Linux on the desktop were to really take off, I think it would change that model.