Making money from open source is a balancing act.
While your underlying product is forged in the white-hot fires of online altruism, the success of your business means striking pleasing postures for the investment community.
As the largest stand-alone Linux distribution, Red Hat has had to perfect this balancing act the hard way.
Commentators were quick to accuse the company of favouring its business arm last November when it opted to split its distribution into a supported enterprise product and a rough-and-ready developer-friendly version called Fedora.
Red Hat claims these kind of changes are a vital part of its strive for profitability -- a sensitive subject after the Linux company, which recently reported the resignation of its chief financial officer, was forced to restate its revenues and earnings for its fiscal years 2002, 2003 and 2004.
Financial dealings aside, the company is also facing increasing pressure from the likes of Novell and Sun, which have shifted their Linux efforts into top gear over the last six months. Not forgetting the fear, uncertainty, and doubt wrought by SCO's legal action and the threat of Microsoft potentially stockpiling a patent arsenal against Linux.
ZDNet UK sat down with Red Hat's European marketing director Paul Salazar to discuss cuddling up to Wall Street while mollifying the Linux faithful -- and whether all the legal FUD directed at Linux can be traced to one source.
How satisfied are you with the status quo that you've established between your open-source roots and profit-focused future?
We are making what we think are sensible decisions. We have had to evolve this business model -- it's not something anyone had done before and we made a lot of screw-ups along the way. It was a messy, ugly process. What we have today is not even the end game -- it is two-thirds of the way there. This model is far from baked.
What we have today works and it allows HP, Dell, IBM, [and] Fujistu Siemens to pre-load technology on their hardware. It gives them something to grapple with. We are in an evolution -- in the early part of the adoption curve.
The market has changed dramatically in the past year, with Novell buying Ximian and SuSE, and Sun firming up its Linux strategy. Are you confident you're being proactive enough after being the front runner for so long?
Have you heard Jonathan Schwartz [Sun's COO] recently? He's controversial to himself. He probably looks in the mirror and says 'I don't agree with you'.
Sun is out there but take everyone else: Novell Suse is an interesting one. They have enterprise Linux business but they've still got the retail market. They kind of still have their finger in that game. Red Hat said: "Look, we are only 650 people globally, trying to do an enterprise business, trying to be reliable and stable to the Oracle's and IBM's of this world -- we can't serve two masters." We said: "There is no way we can do the consumer business -- answering my Mom's phone calls -- and at the same time fixing corporate downtime -- it's inconsistent." We said: "Let's get some focus in what we are doing or we are going to be going in every direction and making no-one happy." That's kind of where we were three years ago. We are not making everyone in the enterprise sector happy yet, but we now are doing the right things and investing in the right sectors.
By splitting off Fedora, you angered some elements of the open-source community. It almost seems as though Novell is doing a better job of marrying the proprietary and open-source worlds…
Red Hat's software technology is 100 percent open source. If you want to put us in a box, then the most obvious box is that Red Hat only, exclusively sells and supports open source -- GPL'd software. I would characterise Novell as a proprietary company that is deploying on an open-source platform and therefore has an open-source strategy.
IBM helped Red Hat get on its feet, but now they see you as a potential threat. What's your response to the claims that IBM invested in Novell/Suse to put competitive pressure on Red Hat?
SuSE did extremely well on all the IBM platforms: they are in on the S390, they are in on the Zseries, they're in on the P-series. OK, how much money do they make from those lines? Almost nothing, right? IBM loves it. They have got SuSE doing their Linux development. Red Hat does some, but we made some strategic decisions and only do that where we think its right for open-source technology. There is no such thing as a free lunch in this case. The price IBM paid was $50m. Think of it as a long-term subsidy for all the work that had come before. They had to feed that into the programme to keep it breathing because they like having choice. Who doesn't?
Then comes the other complaint: we are trying to become the Microsoft of Linux. We can never, ever be that. Our model means that no one will ever extort the kind of money Microsoft has been able to get. It just can't happen, it's just not possible. There won't ever be another Microsoft, and there definitely won't be a Microsoft of Linux.
We believe, genuinely, that the stuff we develop, the way we engage all of these open-source projects, is part of a long-term plan. We know that any changes we make will be part of the long-term kernel development process. In a year, there is going to be a new version of the kernel coming out, and all of our features and changes will be in it. SuSE, like it or not, makes decisions that help to drive the IBM platform, which is great. But the risk is that you build custom patches and custom changes that go into their own development chains, but they will never make it back into the main Linux. So in two, three years they will still be supporting what I call dead-end patches at a great expense. It is not a good, sustainable model, in my opinion. They are very smart capable people, but we have a different approach.
How do you balance the competing pressures of the market and the open-source community -- both with their own agendas, but both vital to your business model?
People in Red Hat value freedom and courage. They take risks. They value freedom as a principle. Quite frankly, if tomorrow we decided to release a closed piece of software, literally half the engineers would quit on principal that day. But at the same time, as an organisation and as a business, we embrace accountability. We can provide that balance that says, at the same time someone can be free to voice their opinions and make choices about technology, but we have to be accountable to customers. Shareholders have invested boatloads of money in Red Hat and expect performance and payback. But it is a balance.
A lot of our engineers wear Birkenstocks to work, and their favourite T-shirt each day of the week and, on other hand, they like to get paid. They respect the fact that we do sell, and we have lots of constituencies that buy from us. We have to serve our customers' needs, and if we don't do that well, we go bankrupt. It is an interesting balance.
How big a threat are the 238 patents out there for the future development of Linux?
We have a patent policy that says: we don't agree with patents. As a company we profess that and it's on our Web site. As an organisation, we are opposed to them, but on the other hand we have acquired patents through the 13 or 14 companies we have acquired. As far as those patents go, our position is that if you want to use them in software that is developed using the open-source model, then fine, you can use them without royalty. If they want to close it up and put it into a restricted proprietary business, then we aren't going to allow that. Bottom line: we are not going to propagate our patents any further.
That's our business philosophy on patents. Now, we also happen to use Linux, which at the moment is facing legal patent challenges from SCO and claims that it might even violate patents that are held by the likes of Microsoft and IBM. What do we do about that?
Do you think it's interesting that as the SCO case has begun to lose momentum, the issue of these other patents has suddenly come to the fore?
Isn't that interesting? There's always a fire engine to chase isn't there? It's a little frightening, but there it is.
On that front, we are a technology company, we are not a legal company. But we know that customers are concerned, so we have built something called the open-source assurance programme. The programme states if any code is found to be infringing anyone else's IP, we will take three steps: we will find out if anyone actually has a problem, will ask if we can hire the rights at some point, and if all else fails we will simply rewrite the code. We have a lot of good engineers, so we are confident we can replace any piece of code that we have to. So our solution is a belief in an engineering answer to a legal problem. We do, however, have a small legal fund when it's needed.
Are there really, genuine patent issues with Linux -- or is this all FUD that can be traced back to one place?
I will not promote any conspiracy theories, but you never know, right? We have no knowledge of that.
Munich claimed that its decision to temporarily suspend its migration of 14,000 desktops to Linux was down to fears over outstanding patents. What's your opinion?
It was on, then it was off, now it's back live again. But there are some real genuine concerns around patents.
Gartner claims it wasn't about patents, but about Munich reconsidering the TCO of Linux.
Our friends at Microsoft have done us a huge favour with their advertising campaign. If you ask me, they are fighting last year's battle. They are out there saying 'oh, yeah, that Linux isn't free -- look at the TCO, we can find ways you can actually save money using Windows'. No. 1, if they are making such a big deal about it, it must be important, right? And No. 2, they are helping to sell the message [that] we have to tell people, the thing we have to say to people every time we start; Linux is not free. If your time is worth nothing, then Linux is free, but most customers tell us that time is money and that they don't want to be OS experts.