The Open Road

When an open-source project is working optimally, can proprietary-software companies hope to compete?

Eat my dust, proprietary sloths

Greg Kroah-Hartman, a prominent Linux kernel developer and Novell fellow, suggests that the answer is no. Speaking to the How Software Is Built blog, Kroah-Hartman makes the case that the pace of Linux development leaves competition in the dust:

[The Linux kernel development team adds] 11,000 lines, remove[s] 5,500 lines, and modif[ies] 2,200 lines [of code] every single day.

People ask whether we can keep that up, and I have to tell you that every single year, I say there's no way we can go any faster than this. And then we do. We keep growing, and I don't see that slowing down at all anywhere.

I mean, the giant server guys love us, the embedded guys love us, and there are entire processor families that only run Linux, so they rely on us. The fact that we're out there everywhere in the world these days is actually pretty scary, from an engineering standpoint. And even at that rate of change, we maintain a stable kernel.

It's something that no one company can keep up with. It would actually be impossible at this point to create an operating system to compete against us. You can't sustain that rate of change on your own.

Microsoft might beg to differ, as would Apple, but the reality is that neither is updated as often or as extensively as Linux is, which supports a far broader hardware portfolio than any other operating system in existence.

Linux is pretty incredible. But it's not alone. Mozilla Firefox, Eclipse, and other projects produce best-in-class software at an almost frightening pace.

Can anyone compete with an open-source project at the top of its game?

The answer might well be no, as the top open-source projects are collaborative efforts between multiple companies that pool resources and expertise to drive development. And while it might seem reasonable that a single corporation could best open source's seeming "development by committee" approach, the reality is that well-managed open-source projects have none of the inertia that one might expect from a communal approach.

Quite the opposite.

Having said that, very few open-source projects actually meet the criteria that enable Linux's success. Most appeal to a too-narrow and too-small population of developers (i.e., single-company projects) to glean the benefits and scale of Linux-like development.

As such, the proprietary-software companies probably won't have to worry about competing with indomitable open-source competitors. Not most of the time, anyway.

For those that do, however, better stock up on the pumpkin pie. It may be the only thing to be grateful for this Thanksgiving season.

Greg Kroah-Hartman interview discovered via @glynmoody's ComputerWorld blog.

For those new to open source, whether on the business or development side, it's hard to appreciate just how far the movement has come in the past few years.

In 1998, when I had my first taste of open-source software through my company's investment in Cobalt Networks, virtually no one knew what open source was, including now-common projects like Linux. Things were a little better in 2000, when I joined a Linux start-up (Lineo), but I spent much of my time working with prospective customers to ease their concerns over open-source licenses like the GPL.

The world is open source's oyster.

By 2004, when a group of friends and I founded the Open Source Business Conference, there was significant, growing awareness of open source, but its adoption was still stymied by Fear, Uncertainty, and Doubt, much of it fomented by Microsoft (Steve Ballmer in 2001: "Linux is a cancer") and the SCO Group (lawsuit over the provenance of Linux code in 2003).

Today, SCO Group, once a high-flier, is struggling for existence. Meanwhile, Microsoft has committed another $100,000 to Apache Software Foundation, has started its own open-source foundation, and has embedded significant bits of open-source code within its proprietary programs, among other things.

Linux, for its part, struggled to get noticed in data centers back in 2003. It has since become essential, mission-critical infrastructure across the Global 2000 ranking of public companies

We've come a long way.

This progress reflects itself in the job market, where Linux-related jobs have seen a 6 percent rise in 2009 alone, while Windows-related jobs have plunged by 8 percent, according to data from Dice.com.

But it's also evident in enterprises' willingness--even eagerness--to discuss open-source adoption plans. Virgin America CIO Ravi Simhambhatla tells The Register that his need to do more with less drove the company to adopt open source and suggests that the open-source philosophy is a positive, disruptive force:

Our company doesn't need just another IT team, the more and more we get entrenched in the...way of doing things the less and less room we will make for ourselves to be innovative.

In 2004, when I was trying to find an IT executive to speak at OSBC, it was a lost cause. No one wanted to paint a legal bull's-eye on themselves for SCO or Microsoft. Today, company executives line up to talk up how they're differentiating through open source.

Open source has "arrived," and the signs are everywhere, from the U.S. Defense Department's efforts to boost its open-source adoption further to patent-rich Qualcomm's foray into open source.

Open source is no longer a question of "why" but rather one of "how." It's the way the industry does business, and the way it does development.

No, not everyone in the industry, all of the time. But for those of us who have been involved in open source for even the past five years, it's amazing to see how much things have changed, which suggests they'll evolve even further.

For some within the open-source world, this is unwelcome news. They defined themselves as freedom fighters, battling the forces of proprietary darkness. And as far as good-and-evil metaphors work in technology, they were.

But as that world embraces open source, they're largely left bereft of bogeymen, like old soldiers still struggling against an unseen enemy.

Winning can be a bit disorienting.

All the same, it's time to move on. There are no more vampires to slay, but simply further open-source education to undertake. Enterprises need open source now, more than ever, and they're adopting it now, more than ever.

What a long, strange trip it's been.

Linux jobs in the United States are booming, up 6 percent since January, according to data from Dice.com. This will come as small consolation to Novell employees, however, which weathered another round of layoffs at the Waltham, Mass.-based company.

According to several sources within the company, and confirmed by Novell's public-relations director, Ian Bruce, Novell last week laid off 100 to 130 people of its roughly 3,900 global employees.

While my sources indicated that the Workgroup division was particularly hard-hit, Bruce told me that the cuts came "across the company, both geographically and productwise."

Novell appears to be doing its best in caring for these employees, offering several months of severance pay, apparently based on the number of years with the company, among other factors.

For those remaining employed there, Novell announced this week that it would be suspending 401(k) matching contributions, which followed on the heels of its formal filing on Monday, to that effect, with the U.S. Securities and Exchange Commission.

Novell has spent the past few years attempting to reinvent itself as a Linux company, and it has managed to string together several quarters with strong earnings in its Linux business on the back of its controversial partnership with Microsoft. The company has struggled to compete effectively with Linux-leader Red Hat.

On November 2, a Novell PR representative contacted me to arrange a conversation with CEO Ron Hovsepian about Novell's "new focus in its strategic direction."

Whether this means more or less open source is not yet clear. It is clear, however, that Novell needs to focus more on top-line revenue growth, and not merely ways to cut costs. Until Novell learns to grow business, and not simply reduce expenses, its employees are going to remain all-too-familiar with layoffs.

The day open source became big business is the day that open-source development exploded. Yes, open source predates the moneyed interests hankering to use it to competitive advantage, but it really wasn't until IBM dropped $1 billion on Linux that companies began paying employees to write free software that the movement saw broad adoption.

That's when open source became more than an efficient way to develop software, and also became a great way to build a business.

However, adding open source to one's business is not magical pixie dust that guarantees its viability. As IBM's Bob Sutor explains:

The basic principles around revenue, profit, loss, taxes, payroll, overhead, accounting, sales, incorporation, health care, and human resources all apply. You can be a starving open source software entrepreneur as easily as a starving proprietary software entrepreneur. No one will excuse basic business failures and screw-ups just because you use open source. Make sure that you will produce a product that people want and in some way will pay for, no matter how indirectly.

Sutor's counsel applies to any company or individual that wants to build a business around open-source software, but arguably some of the industry's best projects are not the product of any one company, but rather of several. Linux, Mozilla, Apache Software Foundation, Eclipse, and other collaborative communities represent an interesting way to use open source to competitive advantage.

In many ways, open source has become a critical component of the software industry because the market has largely moved from vertical businesses (i.e., companies controlling all aspects of production, distribution, etc.) to horizontal markets (i.e., companies focusing on their core competencies and depending on others for complementary functions).

Linux: Peace, love, & squeezing Microsoft

But this business control system has a inherent risk. Should an organization monopolize a specific segment of a value chain system they can extract a higher percentage of its total proceeds. If the product, or service, in question is price elastic than those additional proceeds will come from other participants in the value chain system.

Case in point? Windows. By owning the operating system, Microsoft threw a wrench into the collective cogs of horizontally oriented software firms like Intel, IBM, and others.

The industry's response--Linux--is a classic example of the open-source approach to mitigating individual choke holds within an industry, as Prentice goes on to write:

What then does a CEO do when facing a squeeze on their profits because a direct, or downstream, supplier is dominating a segment of the value chain system? Besides negotiating a better deal - if they can - they've been left with little choice but to get directly into that segment of the value chain system themselves. But by doing so their organization is distracted from focusing on its own core competency.

The risk of such an undertaking can be mitigated if there is a collective response by similarly affected members of the value chain system. After all, it is usually a shared problem. But collective responses have always had an inherent, and often fatal, flaw. Who owns the resulting assets? Either organizations enter into complex joint venture agreements to sort this out or run the risk of shifting the distortion in the value chain system to another organization.

Again, Linux offers the perfect example. IBM, Intel, Red Hat, and others aren't investing in Linux because they're all chums at the country club together, but rather because they're looking for ways to reduce Microsoft's hold on their own businesses through its control of personal computer and server operating systems.

As an added benefit, it's a great way for companies to collaborate without running afoul of antitrust laws. It's collusion without the collusion.

Intriguingly, even Microsoft is getting into this game. Microsoft's partnership with open-source ad serving company OpenX indicates that Microsoft, too, is figuring out how to use open-source complements to loosen strangleholds competitors like Google may be hoping to throw in its way.

This is why open source is growing so much faster than the rest of the industry, as IDC finds. It's not because we love each other more. Quite the opposite. It's because proprietary vendors have figured out that open-sourcing key complements to their core businesses can be strategically decisive in hurting competitors while helping themselves.

Apparently, Qualcomm didn't get the memo. Open-source developers as a group tend to be hostile to patents, believing that they're detrimental to technology innovation.

But Qualcomm, a company devoted more than most to acquiring and prosecuting patents, announced Monday the launch of a wholly owned subsidiary called the Qualcomm Innovation Center (QuIC) to focus on open-source development for mobile.

The patent king seeks to become the open-source king?

Maybe. Maybe not. The mission of QuIC signals an intent to blend the best of mobile open source with the best of Qualcomm's proprietary technology:

Open source and community-driven software development is becoming increasingly important to the wireless industry....Qualcomm Innovation Center, Inc. is a wholly owned subsidiary of Qualcomm that brings together a dedicated group of engineers focused on this area of growing innovation. With the goal of investing greater resources into enabling and optimizing open source software with Qualcomm technology, Qualcomm Innovation Center, Inc. works closely with the open source community to enable the faster advancement of the wireless industry as a whole.

It's a welcome sign, but as yet Qualcomm has demonstrated negligible involvement with any open-source community One of the cardinal rules for engaging with open-source development communities is to, well, engage with them.

Typically, this means writing and contributing code. Code is the coin of the open-source realm, and I'm unaware of much involvement from Qualcomm in this area.

So let me offer a suggested shortcut for Qualcomm: hire someone to educate you. Danese Cooper, formerly of Intel and Sun Microsystems and recently departed from Revolution Computing, could help to shake things up on Qualcomm's San Diego campus. And there are others.

In whichever way Qualcomm opts to do it, the company must engage with open-source communities through free code transfer in order to be taken seriously and to have a chance of influencing such communities.

For example, as noted in GigaOM, Qualcomm intends for QuIC to help it optimize its technology for Android, Chrome, Moblin, and other mobile open-source projects. Yet given that the company doesn't even show up in the list of top Linux contributors, how can we expect to see Qualcomm play a meaningful role in distributions like Android?

No one is going to give Qualcomm bonus points for creating a subsidiary to focus on open-source development, if little open-source code is actually contributed.

In this Qualcomm could learn a lesson from Adobe Systems, which despite maintaining a healthy business in proprietary software, is learning to engage productively with open-source communities, as highlighted in The H Online.

In sum, it is welcome that Qualcomm finally sees enlightened self-interest in leveraging open-source software for the good of its business. Now it just needs to learn to accelerate such benefits through real code contributions--and not simply nice-sounding business units.

It takes time, leadership, and a fair amount of luck to successfully build an open-source community. It also takes money. Lots of it, if IBM's $1 billion commitment to Linux is any indication.

Unfortunately, the return on such open-source community investments may be permanently scuppered by the European Commission's misguided defense of MySQL from Oracle's intended acquisition. If the EC is going to punish successful open-source endeavors like MySQL, will investors still clamor to finance the rise of open source?

In many ways, MySQL is the quintessential commercial open-source success story. On the financial side, MySQL managed to build a vibrant business, doing north of $90 million at the time of its acquisition by Sun Microsystems in February 2008.

Equally compelling, however, is the exceptional user and developer community that formed around the open-source database project, registering tens of millions of downloads and a massive developer community.

This community augmented MySQL's financial fortunes, of course, but it also protected MySQL database users from the whims of the company, as former MySQL CEO Marten Mickos wrote to European Competition Commissioner Neelie Kroes:

Even if Oracle for whatever reason would have malicious or ignorant intent regarding MySQL (not that I think so), the positive and massive influence MySQL has on the DBMS market cannot be controlled by a single entity - not even by the owner of the MySQL assets. The users of MySQL exert a more powerful influence in the market than the owner does.

Unfortunately, the EC seems intent on punishing MySQL--both community and company--for its success. Already the MySQL database project has started to fracture into competing forks, while business rivals like EnterpriseDB and IBM collect confused customers.

More worryingly, the EC's actions may end up diminishing potential returns to investors in other open-source projects, particularly those that take the added time and cost to build global communities.

Technology mergers and acquisitions activity is at a 20-month high. Open-source companies, however, may miss out on this resurgence, particularly those, like Acquia and EnterpriseDB, that build on successful open-source communities (Drupal and Postgres, respectively).

Indeed, based on the EC's actions, perhaps the worst thing these companies could do is foster successful open-source communities. Maybe they should just take the cash and run?

Consider: the EC didn't challenge Yahoo's acquisition of Zimbra, VMware's acquisition of SpringSource, Citrix's acquisition of XenSource, etc. What do they have in common? Rising revenue but, except in the case of SpringSource, much more limited communities than MySQL. (Even the Spring community pales in comparison to MySQL, impressive though it is.)

Granted, the major difference with Oracle/MySQL is that the two are ostensibly competitors, as CNET points out. In the letter referenced above, however, Mickos dismisses such competition. The reality is that MySQL and Oracle compete in two different database markets.

Regardless, as well as MySQL was doing, $90-plus million is spare change in the global database market. The EC, in other words, isn't trying to protect MySQL's business. It's trying to protect MySQL's community.

Such mollycoddling of an open-source community is destructive to all future investments in similar endeavors. Why should commercial entities bother fostering community--the very community that makes them less susceptible to hostile takeover and anticompetitive forces--if doing so simply ends up ruining financial returns?

The EC means well, but it is not doing the right thing for MySQL, its community, or other open-source commercial efforts. Quite the opposite. Just as the commercial open-source community has been pondering a move back to community-controlled open source, the EC threatens to hobble the shift.

The EC may well end up with less competition, not more, by blocking Oracle's proposed acquisition of Sun and its crown jewel, MySQL.

Novell has been positioning itself as the Avis of Linux, a distant but gaining Red Hat competitor that "tries harder." Like Oracle, Novell argues that it can give customers Red Hat value at a lower price.

What, me worry?

It's true that adoption of unpaid Linux like CentOS is booming, and that this no-cost alternative to more expensive solutions like Red Hat is a real threat to Red Hat. This is no doubt why Red Hat has made "free-to-paid" a core element of its ongoing strategy, as related in its recent earnings call.

But it's a much bigger threat to Novell and Oracle, both of whom are trying to position themselves as cheaper alternatives to Red Hat Enterprise Linux.

If a customer really wants Red Hat at a lower price, they're not going to move to an incompatible distribution that may or may not run their applications properly. They're going to jump to CentOS, which is basically a carbon copy of RHEL, minus the trademarks (and price tag).

Oracle, for its part, is clearly not in the Linux market. It's in the market to eradicate Red Hat, so as to claim top-to-bottom control of its software stack. But even as Oracle tries to squeeze Red Hat into oblivion, CentOS provides an excellent hedge against commercial competition from Oracle (and Novell), making its pitch ring hollow.

CentOS: Red Hat's biggest annoyance and greatest friend?

It's not dissimilar to the role that piracy plays for protecting Microsoft's Windows dominance against Linux, especially in emerging markets. Quite possibly the worst thing that Microsoft could do, as IDC has also suggested, is to succeed in its anti-piracy efforts.

Were Microsoft to raise its pricing above $0.00 in such markets, suddenly Linux would look like a much better alternative.

Back to Novell and Oracle. It's not enough to try harder. Red Hat has created a dominant global brand that CIOs trust. It's not worth a few dollars here and there to disrupt that to shift to SUSE or Oracle Enterprise Linux.

Not when those CIOs can shave 100 percent of their RHEL subscription costs by moving to CentOS.

I know some CIOs who have, but they tend to be enterprises with lots of developers that are comfortable supporting themselves. Fortunately for Red Hat, few CIOs care to take that risk. Unfortunately for Novell and Oracle, those who do want to save all of their Linux subscription fees, not just some of them.

In the battle for supremacy among the software industry's Big Four, Cisco may be placing the biggest bets and angling for the biggest returns. Some still think of Cisco as a networking hardware vendor, but hardware is simply Cisco's beachhead into others' turf, similar to how Microsoft (desktop), Oracle (database), and IBM (everything) are using core strengths to move into adjacent markets.

If anyone needed further confirmation of Cisco's software aspirations, its forays into Linux offer a strong hint.

In what might have looked like a publicity stunt around a $100,000 prize for Linux developers, Cisco's Linux development contest was actually a major clue as to just how serious it is about becoming a leading server vendor with a global development community--and soon.

Today, Cisco announced the winners of its "Think Inside the Box" contest. The three winning applications are very interesting, but the bigger story here is what Cisco's contest just demonstrated:

Most of Cisco's 7 million installed Integrated Services Routers (ISRs) are now servers, for all intents and purposes.

The contest proved that server-side Linux developers who know C/C++, Java, or Python can now write applications to Cisco routers with little or no knowledge of routers. (Remember: the finalists only had 90 days to write their applications).

That's a development community of millions, folks. Overnight.

Still think Cisco is a hardware company? By fostering a developer ecosystem around its core router family of products, Cisco just made its hardware solutions much more valuable to its customers (and increased the stickiness of its customer relationships), and turned its routers into a big target development platform for developers.

I wrote about Cisco's contest last June as Cisco's way of paying developers to stick a finger in the Microsoft eye with a $100,000 bounty for writing Linux-based applications for its AXP (Application Extension Platform).

I clearly underestimated Cisco's ambitions.

This is doubly clear when correlated with another Cisco announcement this week about its new and expanded Cisco Developer Network, which SearchNetworking covered.

Cisco is serious about software and fostering a global developer community. As I argued in my "Software's Big Four" blog, each of these companies is entering new markets from incumbent positions of strength, unlike HP and SAP (which both have big software businesses), which are largely sticking to existing businesses.

Millions of Cisco routers already sit in data centers and branch offices around the world. They consume less power than servers. They have a smaller footprint. They're more secure. And they enable a class of applications that Cisco calls "network-aware." Just slot in an AXP blade hosting an application.

Basically routers are much smarter now, and with the right applications can be used to take control of your phones at night to monitor for burglars; manage HVAC, water, and power in your office; deliver advertising in your retail store; and much, much more.

There are two things Cisco still lacks, however, in order to make an unimpeachable bid for developers. First, it needs to move off Broadcom chips for its ISRs and add x86 chips to the mix, something that I'm hearing rumblings may well be on the way.

Second, as impressive as Cisco's outreach to Linux developers has been, the company also needs to support Microsoft's .Net/Windows developers. It's too big a market to ignore.

If Cisco can deliver on x86 and to Microsoft developers--and I think it just might--Cisco will have opened its router (server) family to an even larger development community than the already large Linux market, further blurring the distinction between routers and general-purpose servers.

The result? A formidable software company that sprouted out of a dominant hardware company. How would Oracle, Microsoft, and IBM react?

Follow me on Twitter @mjasay.

Open source is used to playing underdog to incumbent proprietary vendors. What will happen when open source dominates, rather than commoditizes, markets?

I ask because several open-source projects are not far from owning dominant market share in their respective markets. Mozilla's Asa Dotzler reports that Firefox is "on track to easily reach 25 percent of global usage by the end of the year." That may not sound like much, but given that Microsoft has been losing five percentage points of browser market share each year while Firefox gains five percentage points, and it's not hard to imagine Firefox surpassing IE's market share by early 2013.

Linux, for its part, is still only 13.8 percent of the paid server market, while Windows Server still claims 38.1 percent market share, according to IDC. It has a long way to go, but in some markets like cloud computing and the growing Web 2.0 market, it plays a more authoritative role.

So, what happens when these and other open-source projects dominate their respective markets? Will it change how we market open source? Will it mean more research and development dollars must be invested?

Traditionally, open source has done a fantastic job of commoditizing expensive, well-understood markets. While I believe open source can innovate, particularly with companies behind open-source projects, it's still an open question as to whether the financial returns from open-source sales can pay for the heavy R&D and marketing costs that are generally required to create new products and new markets.

Open source has been better at business-model innovation than product innovation, though there are some notable exceptions.

Forget innovation for a minute, however: what will we do when Microsoft, Oracle, etc. are the runners-up, not the market leaders? Microsoft is a convenient (if inaccurate) proxy for all things that are bad in the software world for open sourcerors, but imagine the shift in thinking required to compete when, for example, Firefox has 80 percent market share and IE owns less than 20 percent. Who will we blame for our problems when our straw men are gone?

Perhaps none of this matters, however, as we could see dominant community-led open-source projects fork themselves long before they reach critical, market-dominating mass. It's not hard to imagine splinter groups forming within big open-source projects to take them in different directions, even as Joomla did with Mambo, Ubuntu did with Debian, etc.

The antidote to this is the open-source foundation. Among the examples of strong open-source projects that haven't forked--Eclipse, Apache Web Server, Mozilla Firefox--foundations have been critical to keeping these together. Linux, for its part, has been forked many times, but its core is held together by the Linux Foundation.

I believe the key to attaining dominant market share, and to preventing forks, is the open-source foundation. Over time, I suspect we'll see more "open-source companies" separate themselves into foundations, to manage the code, and corporations, to manage the monetization. This may be the only way to both liberate and dominate at the same time.