The Good, the Bad, and the Ugly

Mark Driver
Mark Driver

Gartner predicts that by 2011 at least 80 percent of all commercial software solutions will include substantial open source. Could you explain the risks and rewards?

Driver: We’re finding that there are different business models for open source. There’s a sort of non-profit model where vendors and various developers support an open-source project. One way they monetize it is through a direct revenue strategy where they’re selling open source. The classic example would be Red Hat.

But the model that we think is going to dwarf any direct-revenue space is for vendors to embed open source inside other products. This includes both hardware and software, leveraging open source as a building block.
The vast majority of commercial software vendors are going to learn that “I don’t want to reinvent the wheel” – I don’t want to spend all my engineering budget on plumbing, on pieces of my software package that are not a competitive advantage. The smart vendors are also learning that they can leverage open-source modules, whether a small part or a huge part of what they’re delivering. Over time, you end up with 80 percent of all commercial software in one way or another leveraging open source inside it.

You called open source “the biggest change in corporate IT since distributed computing.” Could you tell us why?

Driver: Don’t reinvent the wheel. Instead, focus on real value-adding efforts and leverage open source as a commoditized foundation. From a commercial-vendor and end-user-technology perspective, those companies that get it right are going to enjoy significant benefits from having access to the universal pool of free intellectual property and technology.

The vendors that get it wrong are going to end up tilting at windmills and competing against commoditization. That approach never wins. You can’t sell something when there is no value added over what others are giving away for free.
Companies were finding that a combination of Intel and Linux gave them 80 percent of the functionality they could get from SPARC and Solaris, at a fraction of the cost.

Open-source software is now in its “third wave.” Could you talk about what the three waves involve?

Driver: The first wave of open source involved evangelism by the fringe elements and the elites of the sub-culture of the IT industry. Then, we began to see open source making its way into the business model and really being evangelized.

Next, it was going to revolutionize the world and drive empires like Microsoft and IBM out of business. Of course, that hasn’t happened.
What’s happening today is a much more pragmatic approach. It’s less about revolution than evolution. It’s less about taking a dogmatic position on open source and more about looking for real value – where it makes more sense – and using it in combination, maybe still with Microsoft or IBM. That was the third wave.
The first wave was the secret, and the second was about the revolution – which didn’t happen – and the third wave is about pragmatic leverage.

Open source is now mainstream. Could you discuss some of the likely changes in adoption patterns through 2012?

Driver: One thing we’re seeing today is that open source is across markets and across adopters: No matter what the profile, open source is being leveraged for increasingly mission- and business-critical applications. Many core, system-level services – ranging from operating systems to middleware and Web servers – run on Linux. True “business level” processes are just beginning to emerge among enterprise applications.

The other trend is that, as open source makes its way deeper into the market, it’s being introduced into increasingly conservative IT organizations. It’s no longer the bleeding-edge technology adopters; it’s people in the middle of the bell curve.
The majority of open-source users used to be technology-aggressive adopters, focusing on factors such as flexibility, vendor independence, time to market, and standards compliance. Today, open source is making its way into higher-profile, more mission-critical solutions. Adoption is going beyond simple, direct use as an increasing number of technology providers embed open source in their own closed-source solutions. Increasingly, risk-adverse adopters are looking to open source to meet potential IT challenges.
Cost and risk were there before, but they were lower on the priority layer. What’s happening today – and why we’re seeing a significant shift in buying patterns – is that when you start leveraging open source for mission-critical solutions, combined with the increasingly conservative overall profile of the adopter, the demand flips upside down. All of a sudden, cost and risk are more important than freedom and flexibility.

How will the open-source software community expand and divide? Could you provide any scenarios for how this would play out?

Driver: Related to what I’ve already described, some projects that are garnering significant third-party commercial support have processes in place that control their intellectual property.

You’re going to find that there will be two classes of open source: a traditional community-oriented open source, and what we’re tentatively calling business-class open source. Only a subset of all open-source projects will be capable of addressing the needs of those more conservative users. We will see a split – perhaps only a minor split, but we will see two classes, or two levels of open source emerge.
There are hundreds, even thousands, of projects on sites like that may not have the level of governance or support needed for more conservative users.

What are some of the factors that should drive customer open-source software decisions?

Driver: Fitness of purpose is one. Open source or not, you shouldn’t use software that doesn’t meet your business needs. It seems obvious, but open source does tend to get a lot of zealotry behind it. You have to be careful about people trying to convince you to use open source just because it’s open source. That’s generally not a good reason to use it.

The second factor is maturity – the maturity of an individual project is going to provide the baseline for the risk associated with that product. A larger community is going to have more vendors behind it, more community members behind it, and it’s going to be more stable.
The third factor is the adopters’ risk profile: Are they early adopters, mainstream adopters, or conservative adopters? Early adopters can often leverage a product that’s less mature and do so safely because they may have their own internal R&D efforts.
The fourth factor is the deployment scenario. Is this a product that has to run with with 999.99 percent uptime. Is it truly mission- or business-critical; if it crashes, is it OK for it to go down for a minute, an hour, a day, a week – sometimes yes and sometimes no.

Could you explain your statement that open source is not quite as good as some of its proponents would have you believe and not as dangerous as some detractors might suggest?

Driver: All I am saying is that you have zealots on both sides. Think of a political party: Not all Democrats are angels, and not all Republicans are evil, for example. It’s the same thing with open source. Open-source proponents typically say that Microsoft has zero value and Linux is perfect. Well, neither piece of that statement is accurate. You have to get away from the techno-religious fervor around open source to dig through to the real value, and you have to do that on both sides.

That’s exactly why this is product is so popular for Gartner, because we do not have a vested interest. Gartner is one of the few outfits that can look at it and say, “We’ll call out the good, the bad, and the ugly.” Our job is to take a measured, unbiased view of it; we’re not invested one way or the other.

What challenges do you envision for open source?

Driver: The big issue is that, as it matures, it’s going to be leveraged for heavier and heavier workloads. They’ll be greater expectations, and more conservative users are going to demand higher levels of service level agreements.

You’re going to find legal protection becoming a legal issue too, because more conservative users are going to demand a warranty indemnity for the code. That has nothing to do with the technology itself, but more to do with protection around potential infringements.
As an example, Microsoft says that open-source products violate 240-some patents. Companies say, “I don’t want to be sued by Microsoft. I need somebody to protect me.” That kind of challenge is going to emerge, and larger vendors will have to get their skin in the game and begin to protect the open-source ecosystem against such potential threats.

Where would you say open source has achieved dramatic success?

Driver: Clearly, the areas where we see the most significant input is the infrastructure markets. In some areas of middleware, the operating system is a perfect example of where open source is already an incumbent leader. Application servers, Web servers, integration brokers, and things that run on the server in the datacenter – these are areas where there’s already significant success.

The world’s most popular Web server, the Apache server; the world’s most popular Java development tools; and some of the most popular Web development tools – PHP and Eclipse – are open source.
The next phase will be office-productivity software – there’s been a lot of talk recently about OpenOffice – and, ultimately, enterprise applications.