In the middle of a rant about clouds a few weeks ago, I made a broad, sweeping and likely misguided statement: "Today's private cloud wars are just like the desktop OS wars of 20 years ago."
My certainty may have been fueled by a glass of wine too many, but this didn't stop me from diving in further. Each dominant private cloud platform has been shaped by its community, its underlying motives and its partnerships -- just as the operating systems of the nineties had been. If this was true, I reasoned, we could learn much from earlier battles.
As I made my case, I came up with the following table on a napkin.
I hesitated to write down this table. It will make a lot of people angry, many of them friends. Some vendors won't like the fact that they aren't on this shortlist. That's partly because the table isn't based on reality -- it's based on the perception that many IT professionals have of the current market landscape, and what many IT buyers tell me behind closed doors.
Recovered from the previous night's festivities, I pulled the crumpled-up napkin from my pocket and thought about it. While the comparisons may be flawed, there are useful lessons here.
-- Integration happens when people rally around one thing. Unix only really took off when it had a center of gravity (Linus Torvalds) and a company that needed to standardize things (Red Hat.) Right now, while there's a lot of visibility, there's little consensus. This undermines the labor-saving value of clouds, because it means we spend a lot of time on integration and custom development.
-- It's not just the desktop. While Microsoft was the dominant desktop provider for decades, the rise of mobile computing and tablets blindsided it. Cloud solutions that don't work across other devices risk succumbing to a similar fate. As I often forget, streaming apps and managing remote desktops are two of the big use cases for clouds.
-- Services shake things up. In an unfortunate one-two punch for Microsoft, the shift from software licenses to SaaS is hard. Thousands of salespeople and channel partners, whose mortgages and moorings depend on commission from licenses, are learning that many clients would prefer services to software.
-- Apps matter more than legacy protocols. IBM spent a lot of time making OS/2 work with legacy mainframe protocols and existing enterprise environments. But eventually, those legacy systems caught up -- you could talk to a mainframe over TCP/IP -- and what mattered was a diversity of applications running atop the OS. Windows won.
-- You can't put a layer on something and call it new. Early versions of Windows were skins atop MS-DOS, but it wasn't until a native, GUI-based operating system emerged that its dominance was cemented. Similarly, clouds are more than just the automation of virtual machines.
-- People are wary of a single source. Apple's insistence on owning the apps, the OS and the hardware turned many technologists off; designers loved their Macs because they "just worked." Citrix has mitigated this with Apache and its Open Core strategy; but it's a tricky balance to strike. Only once the desktop was a given -- and app diversity came from online sites and services -- did Apple's "benevolent dictatorship" of the desktop seem popular once again.
What do you think? Is the private cloud war just the OS war all over again, and are we doomed to repeat the battles of the past? Or is this genuinely new and different, and will today's providers face a separate set of challenges and triumphs we haven't yet uncovered?
This is one of the reasons I love Cloud Connect. It's a great place to have these kinds of discussions -- often over wine and loud voices -- and speculate on where these platforms' roadmaps might lead us.
Alistair Croll will be a keynote speaker at Cloud Connect, taking place Oct. 21-23, 2013. Cloud Connect offers three days of in-depth boot camps, panel discussions and access to a host of industry experts, all designed to help you weigh your cloud options and transform your business. Register for Cloud Connect now.