I received several interesting responses to my post last week, "I love Linux, but it’s not going to save the world," including Marc Wagner's thoughts. Some folks thought I had changed my tune and was now an anti-Linux Windows fanboy whose salary was paid by Dell and Microsoft (not true, by the way). Many, however, were very reasonable, passionate explanations of ways in which Linux really is saving the world, one PC at a time.
Cheap hardware = landfill Labor ends up being cheaper than hardware because we undervalue our natural resources. As the dollar falls foreign electronics get more expensive. I'd also rather not kill the environment that my kids (and your students) are inheriting. Extending the life cycle will become importnat sooner than you think.
Here is an excerpt from a particularly valuable response:
None of them are running on powerful machines, all of the hardware was stuff that was going to be thrown away and that last one is in a mission-critical place (between the kids and the undesirables of the Internet) all day long, every day, on a high-traffic network. Nobody even knows it's there except the IT Manager, myself and the occasional keen-eyed IT Consultant. But it's doing a vital job in a predictable way for no additional expense on obsolete hardware.
And my favorite, owing to my personal biases towards server-centric computing:
LTSP thin clients allow the old machines sans moving parts, that fail, like harddrives, floppies and even cd/dvd drives to be removed. The remaining solid state components are remarkably durable, and off course require less energy to run. Given that machines of even 10 years old are capable of being utilized the younger machines are over kill for the task.
Of course these machines still have 250W+ power supplies, but the increased longevity of use and the reduced demands of hardware still have an impact.
If newer setups included high energy efficiency thin client terminals then those savings could be made from the get go.
Sure, there are some software applications that may require to be run locally on a heavy duty box but they are relatively few., and the 80% 20% law deffinately applies for the majority of applications.
So what does all of this mean? Was I wrong? Yes and no. Regular readers will remember that my cheap Dodge Caravan recently caught on fire. Last Friday, I picked up a replacement, 2004 Honda Pilot. Not the flashiest of vehicles, and certainly not the ultra-manly truck-based SUV that my testosterone-filled brain tells me I really want. However, every time I turn the key, the engine fires up with that low Honda purr, I can haul my entire brood around with reasonable gas mileage, and it pulled out of my snow-covered driveway (a miserable piece of sadism left to me by my excavator) as well as any hulking SUV. I really do have a point here, by the way.
Buy what you need, buy what you can, but overkill has no place in Ed Tech. For me, I was finally able to get regular funding built into our budget with leased equipment that gets replaced on a regular basis. However, I saved huge amounts of money by moving to thin clients and investing more in servers and infrastructure (I also created a model that I can support with fairly limited time). Would every kid love to have a Core 2 Duo in front of them? You bet. Can I justify it for web browsing and PowerPoint? Absolutely not.
The beauty of such a model is that you can very easily turn outdated machines into diskless thin clients, removing the components that tend to fail most. "Real" thin clients tend to come in tiny, low power packages now, offering savings and conservation on all fronts, much like my Honda can carry lots of folks with minimal fuel consumption and very little money spent on upkeep.
The key message here, though, is that we need to balance budgets, conservation, and the human resources required for maintenance and upkeep of aging machines. In many ways, if done right, Linux can allow us to do just that. Thanks for all of the feedback on the last saving the world post.