Those of you who know me will know I’m a bit of a geek. I guess to work in IT you have to be to some extent or another. Back when I was at uni I’d think nothing of spending hours tweaking the voltages of my poor over-clocked Celeron 300a processor to get the last drop of performance out of it. The fact it spent most of it’s time idle didn’t really come into it! When I got a real job and all of a sudden I had to work out ways of deploying and managing thousands of PC’s. Suddenly all that detail started to become less important, indeed it got in the way, simplicity was the key.
Despite all the changes that have happened in the 10 years since then, that same basic rule still stands, and to be honest it always will. Simplicity is always the goal. There’s a great quote from Ray Ozzie about this:
“Complexity kills. Complexity sucks the life out of users, developers and IT. Complexity makes products difficult to plan, build, test and use. Complexity introduces security challenges. Complexity causes administrator frustration.”
I think he’s perfectly articulated the problem. My challenge was that deploying IT out to a company of 18,000 odd people who are scattered all over the world just isn’t a simple thing, there is inevitable complexity there. We just have to abstract it from the end users as best we can, and organise ourselves in a way that manages that residual complexity as efficiently as possible. Wouldn’t it be great though if the solutions we use helped us, if they were designed to remove some of that headache?
To an extent that’s some of the appeal of cloud services. We identify a capability that we want to provide Atkins, but then get someone else to deal with the complexity. In theory all we would need to do it work out a way of consuming that capability. Generally that’s going to be a lot easier than starting from scratch ourselves.
In a way that’s been the focus of the industry for the past 5 years, working out how to deliver complex services over the internet. We’re now seeing truly viable cloud services emerge from this. Salesforce.com is perhaps one of the most successful – and mainstream – business cloud services. Google’s Apps service is doing well, as is Microsoft’s BPOS equivalent (or Office 365 as it’ll soon be known). As well as these complete services there are also interesting cloud services that are designed to be the components of larger systems. MS’s SQL Azure platform can host SQL or Access databases, and Sunguard has an emerging cloud transaction processing offering.
So ‘cloud’ is here, we can – and do – buy services this way. What’s next?
This is where Ray Ozzie comes back in. If you don’t know of Ray, he’s the guy that came up with Lotus Notes. In the 90’s he then developed the Groove collaboration tool, it was way ahead of it’s time but when online collaboration started to feature in people minds MS bought the product and company. He then took over from Bill Gates as MS’s chief software architect. A clever guy then. Just after he joined MS he wrote a memo that basically turned MS’s strategy around and launched it headfirst into the cloud race, it’s worth a read.
Anyway, the quote I mentioned earlier comes from a new memo he’s written. In it he talks about his view of what the next 5 years are going to focus on. Again, it’s an interesting read. He doesn’t say anything particularly revolutionary but he does articulate what we’re already starting to see here very well indeed. In essence he believes it’s all about Continuous Services and Connected Devices.
Ray suggests that internet based services will continue to evolve and form the basis for how capabilities and applications will be delivered, whether that be to individuals or enterprises like ourselves. They’ll be ubiquitous, available to all, and will end up hosting all of our personal and business data. As such they’ll need to be continuously available – downtime would be disastrous. They’ll need to transparently address the security and privacy concerns of individuals, enterprises and governments alike. Sounds a bit like ‘cloud’, but of scale that we’ve yet to see emerge.
With such a huge reliance on these service that we can’t see or touch, how we access them would become hugely important. Of course there will be local software of some sort, whether a browser in the Google model or specific applications as Apple and Microsoft would prefer (a rare area of agreement between them). Ray suggests the difference is that we’ll grow used to accessing these services on devices beyond the computers we use today. There may still be a place for desktop or laptop computers but in time people will adopt more appliance like devices that are cheap, and therefore simple (or dumb) and interchangeable and replaceable. These devices will be use to access and consume these online services, making use of the vast amount of processing and storage they make available.
He also makes the point these devices aren’t necessary just for consuming information, they may also be used to feed information into these services – some connected devices may just feed telemetry or control information. Again we can see this trend emerging today with things like connected home energy monitors that feed information up to apps like Google Powermeter. IBM’s Smarter Planet ideas apply the same principle to instrumenting systems in buildings, the buildings themselves and the rolling up all this data into views of cities and countries.
So in essence the Continuous Services | Connected Devices idea isn’t a million miles away from the world today. Microsoft themselves have long talked about ‘Software + Services’ and ‘Three Screens and a Cloud’ (i.e. Computer, Phone and TV all accessing the same cloud services), a model that is shared with the likes of Apple and their appliance like iPad, and in future by RIM and their PlayBook.
What makes Ray’s memo interesting however is that as well as articulating all of this very well, his position at MS (at least for now) means that this is likely to form a good part of MS’s thinking over the next 5 years. This is a time when we’ll be seeing new versions of Windows and Office appear, and also a time when many organisations will be taking a close look at how they will deliver communication and collaboration. Whether you like MS or loath them, it’s important to understand their thinking as they’ll take a huge part of the IT industry with them, as well of course as some of your clients and customers who also use their software. For now they’re playing catch-up in this field to the likes of Google on the service side and Apple on the device side, it’ll be interesting to see how they fair over the next few years. I imagine we’ll be hearing more about these themes in future.