vnunet.com interview: Red Hat chief executive Matthew Szulik

Red Hat Summit 2006

Red Hat held its second annual Red Hat Summit in Nashville this week. At the event, vnunet.com sat down with the company's chief executive, Matthew Szulik, to talk about the firm's latest initiatives and future directions.

This interview with Matthew Szulik is available as an audio podcast on the Silicon Valley Sleuth blog. 

Red Hat will contribute certification and testing tools to Fedora. Is this signalling an increased development focus within Red Hat?

I wouldn't say increased. I think it's responding to a continued set of needs that we've been hearing from customers and the open source community. Today you saw Red Hat responding to that. With the support of the Fedora board, we'll bring that to market as quickly as we can.

But if you look back at last year's Red Hat Summit, there was a lot of talk about middleware. Now it seems to be more of a developer focus.

There's the combination of the growing success of Fedora with the pending Jboss acquisition, and the opportunity for Red Hat to have the infrastructure and the tools to create a successful developer relationship. Those capabilities are lining up.

In what way?

They are real, there are not slideware. We've talked about the Dogtail project specifically. (You have to love the names that come out of the open source community.)

Dogtail is a project that was developed internally at Red Hat that's all around building a test harness for the certification and testing for software and hardware integration decks. This allows engineers and developers to run tests against a pre-configured set of solutions that they want to build, and look for optimisation, regression testing, etc.

It's our hope that the Fedora board will approve that and that it will find its way into the open source community. In anticipation of the Jboss acquisition closing, these will provide real tools and real technologies for developers to use in an open source model.

Is Red Hat looking to compete with companies like Mercury Interactive?

I think those companies do a good job of what they do and are focusing on. Their business models are very different from ours. We want to focus on that core of open source developers who are looking for a different set of technologies and solutions to build certification and testing around an open source model.

In what way are you different?

Well, as an example, Mercury Interactive and others have a historical footprint in Windows-based technologies for tools and testing. The technologies that we announced this morning will be focused on open source software development.

You announced your intention to acquire Jboss in April. What does that mean for Red Hat's middleware strategy, and especially the Jonas open source application server?

We were very pleased to make that announcement and once we go through all the legal and government regulations, we hope that we'll announce the final closing soon.

What it should mean to our businesses is that customers now can purchase an integrated operating system and middleware solution that is completely open source.

It provides a great platform for application developers to develop in an open source model. As corporations look to develop more open source applications internally, they now have a complete open source stack to build upon. That's something that the marketplace has been requesting for a number of years now.

Jboss is best known as an application server, which Red Hat launched in 2004 with its own Red Hat Application Server. If I am a Red Hat Application Server user, should I be worried now?

We are very impressed by the people at Jboss and the technical capabilities of that organisation. We think they are superior to the Red Hat Jonas implementation. What we would like to do is take the best from the Jonas implementation and make that available to the Jboss developer team where that makes sense.

So we're going to see a merging of the two products?

Yes, but I would use the term 'technology' rather than 'products'.

Sun Microsystems has been really happy with the number of downloads for Solaris, and will be supporting the Ubuntu Linux distribution on its T1000 and T2000 servers. Do you see Red Hat competing more with Solaris and Ubuntu?

The competition has always been there. I couldn't say whether it is more or less. The market will choose and it will be competition as usual.

Some users are less than thrilled by the current Red Hat licensing model, where you charge a fee for every software product, rather than allowing users to pay for support on servers that only they use. Do you see this changing, or is it a fixed part of your business strategy?

When I hear that it's a little confusing to me because we sell a subscription, we don't sell a product. That subscription comes with upgrades, new releases and quite a comprehensive set of services.

As we move into a more heterogeneous computing environment, we are seeing that the complexity of customer's workloads are increasing.

Customers are running on a dual 64 quad-core piece of hardware, for instance, attached to an Oracle 9i Rack Database to an EMC or Network Appliance storage array and there is a kernel fault somewhere or there is a segment fault.

These are really complicated problems to solve, especially if they are done under pressure during peak load periods.

I increasingly hear from our customers that they are getting value from the current relationship with Red Hat and that the pricing is consistent with their expectations. I think the good news about having other solutions is that customers are having choice.

If they don't have that kind of demanding work environment, or they want no service relationship, Fedora is a great implementation. When people refer to the licensing or the flexibility, I think you need to step back and acknowledge the completeness of value that we're delivering under a subscription licence, and not compare it to a one-time product purchase.

This morning you talked a lot about the need for open content and transparency. How can a public company marry capitalistic principles with open principles?

We're doing it, we live it. I don't think that there's a choice. I think if your goal is to do work that continues to move an entire industry forward by your actions, the expectation is transparency.

The expectation is that there are no obstructions or lack of clarity in the technical direction, your governance and compliance, your public financial reporting.

There is no compromise either between the openness of your source code to your financial reporting or to the core values of your company.

We've tried to build a modern global enterprise in a way that allows us to recruit exceptional people to deliver exceptional service to our customers. The expectation of our people is absolute: they want to be open and they want to be transparent and that's not negotiable.

It represents a different management style. It represents a totally different way and a new way of thinking about how you'll live and how you'll build an organisation that is different, not better or good or bad, just different.

But do investors get this sufficiently? Investors love patents that you can put in the books. Red Hat is not a traditional patent-loving company.

Right. Every great sustainable company started out challenging the historical assumptions and what was viewed with great scepticism when they started. They went through that period of cynicism.

After more than 55 per cent of the market gets behind it, then everybody acknowledges it as being a successful approach to a problem.

Intel is an example when it was competing with Fairchild Semiconductor. In the early 1990s, it was Lotus Development competing against Microsoft.

My view is that the people of our company want to continue to challenge and look for new and innovative ways to solve problems. That requires risk and that requires novelty of thinking.

I think what they are expecting out of me as a leader is to be completely open and transparent. I think if we do this the company will continue to succeed for the next 50 to 100 years.

If we don't, we will end up looking just like the other nuts in the business. That's not something we want.