Staying Open

Staying Open

Marten Mickos is living proof that you can take 12 years to graduate from college with an esoteric degree (technical physics) and still become a CEO. Not that he was slacking: Mickos started his first software integration company while in college. He has also sold software in Russia, launched a Linux database for Solid Information Technology, run a telecommunications software fi rm, and headed up an online betting site that went bust during the dotcom crash. Mickos was recruited to lead open-source database provider, MySQL, in 2001 by CTO and cofounder, Michael 'Monty' Widenius, a college classmate. Since then, the MySQL database has become one of the most popular open-source technologies around, used by many of the biggest Web 2.0 sites. And despite openly admitting that it gets paid by only one of every thousand users, MySQL is thriving, to the point that the company is planning an initial public offering.

Do you think the Web 2.0 market is a bubble?
Marten Mickos (MM): It's not like the dotcom bubble. Yes, there is more VC money and more entrepreneurs than is sustainable by the market. But that's how it should be, because then you get this Darwinian system of the fittest surviving. We try to support every one of our customers to be super successful. But statistically, we know that not all of them will be.

You openly acknowledge that only one out of 1000 MySQL users ever pays. Would you ever consider going partially closed-source and making some code and products proprietary?
We've had that debate many times. I think we might win a few new customers, but we would lose 2 million users. We're not ready for that kind of compromise. We also look at other companies who have built closed-source products on top of open-source ones. They don't seem successful.

I think we are well protected against predatory behaviour by our competitors. When you download MySQL, it's just GPL code. But the code is owned by us. We have the copyright, we determine what goes into it, we put in the bug fixes. There's nobody else with that core skill. Secondly, an important part of our paid subscription can't be copied. Our technical support cannot be copied by others. Neither can our monitoring services.

When people ask if Oracle will start supporting MySQL, I say I would welcome it. I actually told them once, 'Why don't you launch Unbreakable MySQL? You can announce it at our user conference, you can buy a platinum sponsorship, I'll mention you in my keynote'. Because I would love to compete with Oracle on our own turf. I would get an endorsement free of charge, and I would get a competitive situation that I easily could win.

I'm not only hoping Oracle will do it; I'm hoping IBM will. I'm hoping Microsoft will do it. What better endorsement is there? And when you have faith in your business model and your core competencies, why fear those things?

Does your growth also depend on invading their turf? Are you pulling users away from Oracle, DB2, SQL Server?
Not that much. If you look at our customer list, most are new companies, growing companies, technology companies, new media companies. Sure, they could have gone with Oracle, Microsoft or IBM, but they said, "This is a new thing; we want the best product. We'll go with MySQL." Those guys never considered anything else. Facebook: Did they ever consider something else? No. It was absolutely obvious their needs were met with the LAMP stack [Linux, the Apache Web server, MySQL and the PHP programming language].

But investors are going to want to know, is there diversification? You have a lot of media companies, but are you getting auto manufacturers, industrial companies?
If that's what they think we have to do, then they are just plain wrong. Some of the most successful businesses today operate in the modern online world, and we do, too. We will welcome the manufacturing companies and all of those once they take a step into the Web architecture. But if they don't, we won't bother going there.

You seem to be a very grounded businessman who just happens to be in the open-source business. Do you get pressure from MySQL fans who want you to be more . . . "lefty" or something?
MM: I am very passionate about open source. And I do believe that it is a superior method. But at the same time, I must be pragmatic. So when they say being dogmatic is very important for the Free Software Foundation - well, they should be. That's what we respect them for. But running a business is not about dogma. We are not judgmental about our customers or partners. We are happy to partner with closed-source companies. We believe that in the long run, open source will win anyhow.

There's nothing leftist about open source. This is a common view that is incorrect. Sharing doesn't make you a leftist. I mean, who are the most sharing people in the world, by charitable contributions? It is Americans.

Our customers are not against us making money; it's that they're for freedom. So make as much money as you can, but don't touch the freedom.

Some analysts cite MySQL as proof that most commercial open-source companies don't rely on outside code contributions - that it's all a myth.
I don't know why people are always glorifying lines of source code. This is a big fallacy in the industry. We are trying to get more outside contributions to the kernel. But building a successful software product is so much more. Source code represents just 10 per cent of the effort. We need contributions anywhere in the value chain. We are happy to produce all the code in the kernel if somebody else writes all the manuals or documents the code or fixes the bugs. The value of that is as big, if not bigger.

Do you actually pay outsiders for contributions or bug fixes?
We've done so many times. It can range from a few thousand dollars to hundreds of thousands of dollars. Sometimes this is to a single brilliant programmer, not a company. We want people to make money in partnership with us. So why wouldn't we pay them?

What's changed at MySQL as you get ready for an IPO?
Of course, there are more procedures and more standards. Our engineers are saying, "Hey, this is becoming more corporate!" But that's the way it goes. What's the alternative? That we don't grow? We have 350 employees now. I think we'll double that in a year or two. I think we all accept what's happening, but we'll still look back and say, "Do you remember in 2002, when the whole company got together and we made those decisions together?"

Follow Us

Join the newsletter!

Error: Please check your email address.
Show Comments