Thin clients provide an incredible number of advantages over traditional PCs:
- They consume very little power
- They are generally silent
- They are easy to secure
- They have little value alone, so tend not to be stolen
- Central management of the server saves a lot of time in client management
- They tend to be less expensive than regular PCs
- Linux and Windows solutions, including virtualized desktops, are mature and plentiful
Frankly, they're just plain easy for even a single IT person to manage; software gets installed in one place, patches are applied to one computer, anti-malware is managed at a single point, etc.
So since they're apparently the best things since sliced bread, why am I considering another computing model as I look at deployments around the district and the next refresh at the high school (about a year away)? A couple of reasons. Thin clients represent a single point of failure: if the server goes down or loses connectivity, entire labs go down. Streaming video and audio can also be problematic, since the server generally have only one or two connections to the Internet and internal network; 20 student pulling multimedia content on what amounts to a single computer from the web can strain bandwidth pretty quickly.
They also don't allow for the sort of collaboration that I'd like to see among students. Although thin client laptops exist, they are quite expensive, meaning that students are confined to a lab setting when using thin clients. They sit next to each other, usually in rows so that they can all see the instructor and will occasionally pull chairs closer together to look at the same screen for small group work.
Regular readers will know that I'm on quite a netbook kick lately. While I'm not suggesting that they represent a solution to every IT problem faced by schools, I have to wonder if a deployment of netbooks might not make more sense than a full thin client solution. They certainly address the single point of failure and collaboration issues. As we move students further into the cloud, as well, the need to access standard Office productivity suites diminishes.
Management obviously becomes more time consuming, although well-configured Linux distributions should be relatively robust. In fact, netbooks could lend themselves to a hybrid approach: use the netbooks for daily computing and a less expensive terminal server (accessed from the netbooks) for a select group of applications to be managed centrally (Geometer's Sketchpad, for example).
I'm certainly nowhere close to making a decision on this; I have plenty of time. However, netbooks seem to represent a nice alternative to expensive mobile thin clients and seem to address some of the deficiencies in the thin client computing model. They definitely warrant consideration, particularly as their prices continue to fall.