Wednesday, July 6, 2011

Nothing but net: stateless computing, technology triggers, and the business of technology

I believe that stateless devices are one of the most important technology trends. The concept is simple and elegant:  all applications and data run from the cloud, with lightweight, fast-booting operating systems delivering a connection to the cloud and a very rich, browser-based user experience.

Devices built as stateless will need much less in the way of complex operating systems, and can be less expensive to build, buy, and operate.

Stateless is important because it changes the discussion about devices in the way social networks change the discussion about communities; we get to manage our lives based on what we need right now, not on being in the presence of the device where the software and data we want is residing.

But why should you accept that stateless computing is A Big Thing? 

Predicting technology is a constantly-shifting balance of art and science. But I do believe we can think of the future in ways that are more roadmap, less crystal ball.  That is the idea behind technology triggers, timelines of changing capabilities in which we can observe A and B and think clearly about the still-unknown C.

Here is a timeline for the enabling technologies of the stateless world:

The vertical lines show the triggers. As an example, the emergence of widely-distributed 3g wireless networks was one of the key enablers for the explosive growth of smartphones and the currently-popular online application markets.  Had you examined mobile computing through this lens just before 3g, and extrapolated to possible outcomes once it became available, you might have gained competitive advantage in the mobile space, just as Apple did with the iPhone.  We are clearly past the trigger point for the so called “app internet”  But we are also close to triggers for stateless computing, with Google’s Chrome OS and Chromebook computers being first to market.

Look at the graph as the timeline progresses; the enabling trends are moving in the direction of stateless computing, which overcomes major limitations in even the most advanced “app internet” situation deployed at the time of this article in summer 2011.  
Further, we can see that a true movement to cloud computing is not complete as long as we must still synchronize cloud and local data and do installation and maintenance of locally-installed applications--a model that is as old as Lotus Notes.   Apple’s vaunted iCloud service, which stores data and music in the cloud, still relies on local copies, a move analysts were quick to remind us, “looks great as far as it goes, which is all the way to about 2006.” 

The “app internet” concept is the future, says Forrester Research CEO George Colony.  I respect Mr. Colony and Forrester, but I think the tech triggers say otherwise.  The use of small-locally installed applications on top of a complex and proprietary operating system, using cloud data sync, is the best we can do in a world where connections are not quite pervasive enough. But it is still kludgy beyond belief, and does very little about the device dependence that now seems very old fashioned

We have not quite reached the “trigger timeline” when stateless takes off, but I think we are close.  What enablers are not quite there yet?

better internet connections: faster everywhere, transparently available via multiple sources IP network providers, especially mobile

quality-of-service management for IP connections

every “-as-a-service” vendor: Software-, Platform-, Infrastructure-

cloud-based media vendors: Pandora, Netflix, Hulu, Amazon
proprietary content delivery architectures: non-IP voice, cable and satellite television services

penetration and evolution of browser-enabled applications HTML 5  

desktop virtualization vendors to stateless-enable legacy applications

app developers

cloud-centric platforms and development environments

interface management systems that adapt stateless UI delivery to different browser forms on different devices.  
applications and developers still relying on client-server models

enterprises with large legacy app portfolios (although see “desktop virtualization” as opportunity to solve this)

rich internet application tools that rely on local installation of software: Adobe Air, Microsoft Silverlight

stateless-enabling operating systems and devices for other form factors: phones, televisions, tablets, automobiles

developers of new web interfaces: voice, gesture, image recognition.

the network effect of stateless benefits across many form factors presents a huge opportunity for manufacturers.  With very fast cycle times for mobile devices, a new market leader could emerge very quickly
fat operating system vendors could quickly be disintermediated: the stateless web experience is backward compatible with legacy operating systems, but not the other way around.

Although there is a high rate of change, there is also a big picture, a timeline from early mainframes, to networked PCs, to the internet-centric, multi-device era.

That multi-device era is where we are now: some cloudsourced data and some applications in the cloud, but still using a client-server model and complex operating systems.  While it's easy--facile, even--to say "the PC era is over," the truth is, we're much closer to a time in which the operating system and the purpose-built device of any sort will become almost invisible, and the idea of "browser" as a separate application becomes irrelevant.  
Stateless is the logical next step, based on evolving technologies, the social use of technology, and the historical evidence of the market in the face of the new capabilities brought on by those technologies.

Stateless is also the enabling technology that unlocks the greatest benefit of cloud computing, and helps us apply the lessons of online social interaction to the way we deploy and use the machines that make it possible. 
It is the tool through which data and applications in the cloud will create explosive growth in business value. Stateless computing is a big idea.

No comments:

Post a Comment