The Future of Open Source
Over the past couple of months, I've started to see open source begin to really take root in IT organizations. Most of them have implemented open source on an ad hoc basis - in other words, outside of the usual IT processes and controls - and seen excellent outcomes: quicker implementation, high-quality support, community sharing, and far lower prices.
As a result, a number of IT organizations I've spoken to are now moving to formalize their open source use; they're integrating it into their everyday processes in order to manage it as an important part of their standard operating environment.
Frankly, it's taken longer than I expected - I always assume people will grasp implications more quickly than they do. It's just the way organizations work, I guess. But having seen the beginnings of this movement, what should we expect the future of open source to look like - what will the IT world look like after open source becomes standard operating procedure?
Here is what I think we'll see:
Islands of Legacy Apps Surrounded by Oceans of Open Source
Nobody is going to yank out what's already in and working - especially if they've spent years and millions finally getting it right - especially given the fact that humans, once having put effort into something, are reluctant to recognize a mistake and move to another alternative. So legacy apps will stay, and legacy vendors will continue to cash their maintenance annuities.
The glory days of proprietary apps are long gone, though. The starkest evidence is Oracle's continued buying spree. Consolidation is a strategy based on improving financial results by removing costs, and Oracle's behavior communicates clearly that they believe all that's left is recarving a fixed-size pie.
However, just because Oracle and its ilk cannot succeed in tomorrow's IT environment doesn't mean that the use of IT has crested: Far from it. As the move to cheaper hardware and virtualization (more on that in a moment) continues, more and more software will be consumed. It will just be cheaper and delivered without the proprietary package trappings. And it will all be based on open source. The result: legacy apps continue into the future with open source engulfing them.
Take This Ticket to the Cash Register and Someone Will Bring Your PC Out From the Back
What will the software industry look like post-engulfment? For a vision of the post-open source software world, a look at the history of the PC is informative. Twenty years ago, PCs were expensive. Stores had experts on staff who offered insight and advice. Then PCs got cheaper - lots cheaper. Today you go to Best Buy and heaven help you if you need anything beyond basic (and I mean really basic) information. The complaints about Dell's service degradation are legion. However, these changes are an obvious result of the fact that there's less money sloshing around the PC channel. Of course, there are many sources of information about PCs available, particularly online. I get a weekly newsletter called Extreme Tech, which has excellent articles and benchmarks - but I still have to be enough of an expert to figure it out for myself.
That's the future of IT. With less money sloshing around the software world, vendors will do less hand-holding. IT shops will need to self-educate about the important technology developments. And, by the way, with less vendor money available, other parts of the technology marketing ecosystem will get less, too. That means less analyst- and PR-generated marketing (I know it gets presented as objective research, but everyone knows it's fueled by vendor dollars).
Consequently, IT organizations will need to develop new sources of information, either internal or external, and will need to pay for it directly. Overall, they'll need to get smarter. One possibility to replace the traditional "free" sources of information (i.e., analysts, etc.) will be for IT organizations to collaborate on research and experimentation as a way to reduce individual costs.
I Build the Foundation, and You Just Drop a Prefab House on It
Today houses are very nearly custom-built from the ground up. Even though contractors work from plans, the house itself is constructed on site - and every house is done as a one-off. Prefabrication is starting to take hold, where the site-specific work is confined to the foundation, and pre-built modules containing walls, wiring, plumbing, and so on are snapped together on top of the foundation. Just as the assembly line vastly reduced the cost of manufacturing an automobile, so too does prefabrication for housing.
The IT analogy of this world is that the future IT organization will create a stable, scalable infrastructure containing common services like network connectivity, identity, and security, and pre-built applications will be dropped onto the infrastructure. Instead of calling these pre-built applications modules, we'll use the fancier term virtual appliances.
Virtual appliances will drop the amount of tedious IT activity by a huge amount. All of the installation and configuration of software components necessary to get an application up and running will vanish since it will all be done by the appliance creator, which is the real expert in installation and configuration of their application.
To echo last month's newsletter, the virtual appliance world will be a world that glories in wasting software - computer processing will be applied in any useful context rather than being rationed by capital and labor availability.
Your Car Dealer Makes More Money on Service than on Selling Automobiles
The average new car dealer, in some sense, uses new automobiles as a loss-leader as a way of forming a service relationship with the customer. The cost of ongoing maintenance serves as a steady revenue stream for the dealer. Overall, car dealers have transformed their businesses, moving from individual high-ticket transactions to a higher-level ongoing value delivery business.
That kind of model is how IT vendors will make money in the future. Rather than selling expensive infrastructure software in one-off transactions, they'll make money by offering services or applications that run on top of the infrastructure. And the value of these services or applications will be far higher to the user.
These new offerings will take two forms. The first will be services to help make the cheap, open source-based infrastructure perform more efficiently. As an IT user, I will pay not for access to the product, but for making the product run well, a goal which can easily be tied to business metrics. In my view, this offering will be well beyond today's typical open source offering of support subscriptions. This offering will be more akin to the vendor taking responsibility for running the infrastructure, rather than merely helping the user in his efforts to run it.
The second form of offering is even more interesting. Because there will be no shortage of software infrastructure, and budget won't be tied up in proprietary infrastructure products, money will be available to fund new software products that run on top of the infrastructure. These applications will be (naturally) delivered as appliances, and will be business benefit-oriented.
Both forms will be money-spinners. I refer to the shift in software business opportunity as "the migration of margin," meaning that infrastructure margin is heading downward (save for legacy implementations, which, as already noted, are high-margin annuities; nevertheless, there is little high-margin new opportunity in the future), but higher-level services will offer great opportunity for margin.
It's a Virtual World
Already mentioned, but worth repeating: virtualization is going to transform both the software industry as well as the world of IT. Installing applications goes from days filled with frustration and swearing to mere minutes. No more trying to figure out cryptic documentation. Up and running. Magic.
It goes much further than that, however. Virtualization offers the ability to shunt virtual machine images around the network, executing them wherever spare machines cycles are available. Virtualization also offers the ability to fail over virtual machines when hardware fails.
In ten years we'll look back on the old labor-intensive mode of software installation like our grandparents looked back on operator-assisted long distance dialing: a wistful nostalgia combine with gratefulness for having moved on. And guess what? Virtualization is going to be driven by open source or by software given away for free, forced to zero price due the competition offered by open source products.
Gee Grandpa, How Bad Were the Old Days?
I've referred to the coming IT world as a time of post-scarcity software. Up to now, we've had to ration our use of software, but open source converts software from a luxury good to a commodity.
We can't even imagine where this will end up, but if you want a peek at software's future, take a look at another mode of publishing: music. That industry is going through a protracted convulsion as digitization makes music universally available. The established order has its hands clenched onto digital music, trying to prevent it from breaking free, all the while overlooking the fact that digital music offers the opportunity to make music available so many more places and in so many more ways that the opportunities are unlimited - but only if the incumbents stop confusing their businesses with the future of music.
Packaged software will go through its own upheaval as established players attempt to keep customers imprisoned in outdated business models. From the outside, this transformation is invigorating; from the inside, it's hellish - but inevitable. The great thing is that you don't have to wish for this future; all you have to do is watchfully wait.
