Open Source Commentary from Navica's CEO, Bernard Golden

In This Issue

• GPL3 Followup

• Musing on Margin Migration

• Navica News

GPL3 Followup

I discussed GPL3 in a newsletter a few months ago. The initial draft of the license has now been released. I am doing a blog on GPL3 for CIO Magazine, discussing the license, its conditions, and how people can submit comments. CIO is just about to move to a new content management system (they're moving to Mambo) and I am scheduled to move to my own blog soon. Depending upon when you try and access my comments, you might have to do a bit of poking around on the site to find them.

Nothing in the license has changed my mind that this will represent the most significant development in open source over the past five years. Once the implications of the license become clear, there will be a torrent of comment and concern. I just can't figure out why we aren't hearing more now. It's something like the Sherlock Holmes story Silver Blaze, in which Holmes remarks that the most puzzling event of the mystery was that the dog didn't bark in the night.

Musing on Margin Migration

I had the opportunity to speak at BayLISA a couple of weeks ago, at a special meeting held at Google. By a fortunate propinquity, I met with MySQL the following day. I was astounded by the extreme contrast between the two companies in terms of facilities, behavior, and, for want of a better word, presence. In my discussions with MySQL, I was surprised to find out that, despite these enormous differences, there is a critical symbiotic relationship between the two. The differences in their circumstances, coupled with the close kinship of their offering, caused me to muse on the role of margin in the technology industry, and where to look for it in the future.

As I said, the contrast between the two companies is extreme and immediately obvious. As I pulled into the Google parking lot, I was met by a security guard who stopped each car to ascertain the occupant's purpose in visiting the company. MySQL, on the other hand, is located in a shared mid-rise building in Cupertino where you park anywhere you please. Google is located in a set of buildings in Mountain View in which they are the only tenant (in fact, the buildings were formerly the home of Silicon Graphics).

At Google, the building lobby was staffed by a number of friendly receptionists, who insisted on registering each of the meeting attendees. We were then walked to the presentation room by one of the receptionists. At MySQL, by contrast, you enter into a bare reception area and find your way back to someone that you can ask to announce you to the person you're meeting.

I was particularly impressed with the spread Google laid out for what was, after all, a low-key community organization meeting. They provided at least five main courses, six salads, many appetizers, barrels of drinks, and even had cake (BayLISA usually makes do with donated pizzas). Backing onto the presentation room was a fully stocked kitchen area containing glass-fronted cooler cases full of drinks, stacks of containers holding snacks (both junk and organic), microwaves, and tables and sofas for relaxation. It even had a full-size pool table for noodling time.

At MySQL I was guided into the kitchenette and invited to select a soda from the cardboard carton that had clearly been brought in from Costco.

After my presentation at BayLISA, a Google employee described the Google infrastructure and product, noting that it runs on “over 10,000” servers (actually rumored to be in the neighborhood of 150,000 servers) using a Google-derived Linux distribution, with a Google-invented file system to contain content and indices.

You get the message. Life is lived large at Google. They're phenomenally profitable. Google is sloshing in so much money that they can pursue the kind of pointless initiatives heretofore seen only in the packaged software industry (Google Pack, the incomprehensible Google Toolbar partnership with Sun, and, just this week, the WiFi hotspot sharing company, FON).

MySQL, on the other hand, clearly watches every penny. No executive chefs. Standard benefits. Low-key marketing. No off-point initiatives: databases is all they do.

The point of all this isn't to describe the extremes of wealth and poverty in the technology industry. At least, that's not the primary point. Google is the global leader in paid search, is publicly held and had about $6 billion in sales over the past year, with a profit margin of approximately 25%. MySQL, an open source relational database company, is privately held, but is rumored to have had trailing twelve month sales of around $40 million with positive profitability. (As a side note, the fact that Google is located in SGI's old buildings provides ironic symbolism. SGI, an exemplar of proprietary software and hardware, has fallen on hard times: no longer listed on Nasdaq, stock at 36 cents, unable to find a way forward in today's technology world).

Despite all that I've just pointed out as the dramatic contrasts between the two companies, the most dramatic thing I learned during my two meetings is that virtually all of Google's paid search runs on MySQL. In other words, the billions of dollars of revenue that Google achieves is balanced on the open source technology of a company less than one percent of its size. That distribution of revenues and profit is so astonishing that it gave me pause: In contrast to the dot com bubble, where Oracle and Sun made billions from Net startups that couldn't find a sensible economic model, Web 2.0 shifts the revenues from the suppliers to the customer-facing entities.

There are a number of lessons in this, and they tell us where the future of the technology industry is going. More critically, they tell us where the future of high-margin technology is going.

Lesson One: Startups Need Open Source

It's easy to overlook today, but Google wasn't always a 500 pound gorilla. In fact, when it started it wasn't even a chimp. It got its start in a dorm room running on a mishmash of hardware (the Google presenter showed a picture of an early setup and it looked like the remnants of a tech garage sale) using open source software. Many people have attested to the fact that open source lets a startup get off the ground with far less investment than used to be necessary -- but when you understand that today's Google was bootstrapped with borrowed hardware and free software, it suddenly becomes clear in a way that dramatically illustrates the point.

Lesson Two: Flexibility Demands Open Source

As previously mentioned, Google runs a home-grown Linux distribution. At least in part, this is so that they can use low-spec commodity hardware in their infrastructure. The Google presenter mentioned that their typical machine is a 386-class box containing a 40 to 80 gig drive (I didn't even know that you could buy that level of hardware today!) Using Linux allows them to run a stripped-down OS sized for their preferred hardware. If they attempted to use a commercial operating system, they couldn't size it for their infrastructure, which would their business model upside down. And, of course, they would be totally unable to replace the native file system with their own version, tuned for their purposes.

Lesson Three: Scale Mandates Open Source

One of the most amazing things about Google is the way they make money. If you've ever used Google AdWords, what's most amazing about them is how cheap they are. Clickthroughs typically costs less than a dollar; many AdWords cost on the order of a nickel per clickthrough..

The only way to build a huge business on transactions of that size is to achieve enormous scale – for Google, on the order of hundreds of millions of searches serving up millions of clickthroughs each and every day. Google's business makes McDonalds's “billions of hamburgers sold” look Lilliputian.

Google efficiently captures the curiosity of the world's Internet users and harnesses it to a very scalable infrastructure. Without the ability to cheaply scale their infrastructure, Google could not convert their transaction model into a high-revenue business -- they literally wouldn't be able to cost-effectively respond to demand. (As a side note, someone from Yahoo was in the audience was at the presentation, and noted many similarities between the two companies' infrastructures; the key difference being Yahoo's use of FreeBSD rather than Linux. Yahoo is also a heavy user of MySQL).

Based on this insight, we can predict that, going forward, Internet-based businesses will have to be based on open source. Without it, these businesses just don't make economic sense. The sole exception will be Microsoft which “eats its own cooking”, relying on Microsoft products to support and run its online properties; one can speculate that its properties do not suffer true intra-company transfer costs; in other words, they get their software for free and so can afford to use a commercial software infrastructure. No other Internet-based businesses will be able to make a go of commercial software-based infrastructures.

Lesson Four: Margin is Moving

Google is a business that couldn't exist without open source. But because open source is available, its founders were able to create a business that addresses the conditions of today's Internet – in other words, respond to the most dynamic, highest-growth part of the globe's economy. It's at the edge of social change, the stress of economic transition, the shift of cultural attention, that high-margin opportunity lives. Because its founders married today's needs to open source, they have realized outsize gains – both organizationally and personally.

MySQL is doing very nicely, thank you. Its stated ambition is to reduce the size of the relational database market by 90% by offering a cheap, easy to use, mid-range product. It will undoubtedly make a very nice living, but will never be a huge company. MySQL will achieve mid-size gains.

The Google/MySQL Takeaway

Several recent newsletters have discussed the impact of margin migration on the software industry and the subsequent impact on IT users. The Google/MySQL story illustrates this story perhaps more vividly than any economic description could possibly achieve. However, the theme running through all these newsletters has remained constant: open source is changing the economic assumptions of the IT industry and forcing old and new players to rethink their strategies. Don't imagine that you are immune to its impact. Get ready to have your world turned upside down. Start working on your plan now – if it's not already too late.

Navica News

You can hear me speak at these upcoming events:

March 17, 8:30 a.m.: SDWest, Santa Clara Convention Center. Presentation topic: "Open Source ROI: Achieving the Promise of Open Source." See www.sdexpo.com for more information.

April 27, 7:00 p.m.: Peninsula Linux Users' Group. Presentation topic: "GPL3 -- Where Free Software is Going." See www.penlug.org for more information.