+

James Wanless

this is where I write

Is cloud computing still blue sky?

As I lurch through this summer trying to finalize the thesis proposal which will frame a large part of my life for the next year, collaborative tools for learning are never far from front of mind with me. A real interest of mine is the use of collaborative writing tools for learning in higher education. In particular, I am interested in learning more about why adoption is so weak – certainly where I work, anyway. Where I get to with my thesis research on that front is another post for another time.

A big clue as to why some post secondary institutions are so jittery about Open Educational Resources (OER) might be contained in the news this past week that Twitter suffered a substantial security breach perpetrated a second time by the same French hacker. An institution’s specific approach to open computing will also have a lot to do with how jittery they should be about it. For example, are we talking about true cloud computing or just using open tools, but hosting them? There’s a big difference from a data ownership and access control perspective in each approach. For the sake of argument, I’ll stick to using the cloud.

Over the past week or so, my feeds were all-a-twitter (and so was Twitter) with news that Microsoft Office’s cloud offering might be imminent. Given the success of Google’s office suite in the cloud, it stands to reason that Microsoft needs to make a me-too play and quickly. Not to be left behind, cloud computing provider Zoho announced a plug-in to migrate their DBs to web apps, either on Zoho or as Google Apps. The big players are pretty certain about the move from the desktop to the cloud. Google’s announcement that Apps could now translate uploaded documents paled in comparison to their bigger one.

Chrome OS is the next Google play in moving everything to the cloud. The idea that you can seamlessly move your desktop work to the browser would still seem to be a ways off, given the weaknesses in some of the Google Apps feature set:

But there are still a variety of applications that simply can’t be replicated within a browser, and consumers have had a mixed reaction to Google’s own apps, embracing Gmail but finding its presentation software to be severely limited compared to its desktop app counterparts.

As Chrome OS will ship with netbooks at first, full desktop replacement at this stage may not actually be on Google’s mind:

The Chrome OS will be appearing first on netbooks, which can’t handle some of the more heavyweight desktop applications in the first place. And the new offering has the potential to drive users to rely on Google’s online offerings, which certainly would further the company’s goals.

This probably spells the end of long-term Android development, as a single, more robust OS from Google to be run on all devices would make the most sense.

While the happenings in cloud computing show the big players are clearly making a push in that direction, moving everyone there will be no small task. For post-secondary institutions, there are many concerns including ownership of data, privacy and, as Twitter just showed us, the risk of a complete breach of authentication data. However, these issues are just the tip of the iceberg.

Technical, SLA and business solvency also pose extreme risks for organizations, let alone individuals, in trusting their data to Google or Microsoft and hosting it in the cloud.

What about power outages?

In the last week alone, there have been several high profile outages at data centers that host sites, such as video site DailyMotion, credit card authorization service Authorize.net, and Microsoft’s Bing Travel. Even the Google App Engine—a platform for third-parties to run their own cloud services—experienced performance issues that resulted in high latency and even data loss.

Or uptime guarantees?

Many large companies are used to having control over and responsibility for all of the servers that the business uses, so the idea of putting parts of their business on rented, “black box”-style cloud services makes them uneasy.

When it comes to education, we often think of OER in terms of access to scholarly journals, tools to help learners and instuctors, social software for social learning activities, repositories of objects, and the value that opening your work up to a wider audience brings. If enough instructors use OER journals, the story goes that more academic publishing will be placed in open, peer reviewed journals. Demand drives innovation and change, after all.

All of this is great in theory and I’m certainly a big proponent of social computing. However, I’d stop short of putting all my eggs in the cloud basket just yet, and maybe ever. I much prefer open source software, but hosting it, where I own the data and support the servers. There’s nothing wrong with using online office suites to help share collaborative work and tools like blogs, wikis, Skype, IM and chat are wonderful for bridging the distance gap. However, I always want to keep my core data, or at the very least backups of everything I’m doing there, locally stored. If I always have to do that anyway, then it starts to dilute the value of cloud computing as a way of providing cheap storage and persistent access.

Frankly, I’d never want to be the one to have to tell a class full of students or anyone else for that matter, that they couldn’t get at that assignment that was due because Google had a server farm meltdown. Or, that institutional data like student numbers, grades, addresses, credit card numbers and the like had just disappeared into a cloud because they were housed there in the first place.

Comments

[…] More here: James Wanless » journal » Is cloud computing still blue sky? […]

Top