I'm not much of a hardware guy. Oh sure I have built a few systems and I can add and replace hardware with the best of them. I've installed hardware since back in the days when installing software meant bringing a tool kit so you could connect wires and cables and all sorts of things that aren't even included in computers today. So I can do it. It's just not something I think of as fun. So I buy my computers "off the shelf" with most additions added at the factory. More on that later.
Now not everyone feels the same way. My friend Max had a great Max Builds a PC series on Channel 8. It was very interesting and he got a lot of great feedback about what parts to use. On paper the system looks wonderful. It's probably pretty good in real life as well. But while it works for one of a kind systems how well does it scale?
Build it your self systems are very tempting as a way to save money. They and their close cousin the "white box" are some of the things that some companies and more and more schools are looking at to save money. Take for example this post by Christopher Dawson. He suggests that by putting his students to work (free labor) and buying parts he can build computers for his school at a much lower cost than he can by buying complete systems from a hardware company. The traditional "white box" vendor follows pretty much the same model but without the free labor. The lack of engineering and other costs (marketing, shipping, etc) lets them sell for less. It all seems so good.
I'm skeptical though. Not because I have any great love for the big hardware companies or because I have a careless attitude about money. No, rather it is that I have had some bad experiences in my short 35 year career and they leave me gun shy. That and because I have worked first hand at a number of companies who design and make computer hardware I have some appreciation for what goes into their boxes. It's not as easy or as simple as it appears. There is some real value in having professionally engineered computer systems.
When I first started teaching high school the school had built a lab out of "white box" computers. They had saved a lot of money. Unfortunately we had a lot of problems with them. Not just the usual early life failures but some of the pieces just didn't seem to work together well. The systems never really became reliable and I was happy to see them go a few years later. The time and effort we (staff) spent and the lost opportunity costs of students not having working computers really seemed to me to exceed the value of the money that (in theory) we saved. I've heard good stories and bad so your mileage may well vary.
Computer hardware is complicated and things do not always work as well together as one might think. How many people build their own Apple computers from scratch parts? Not many. Apple keeps tight control over the computers that are supported under their operating system for some good reasons. First among those is that it allows them to better control the quality of the user experience. PCs don't have that same tight control. The big companies on the other hand do have engineers and testers who do a strong job of making sure the components they use work together. That has added value to me. Perhaps not as much as the cost to everyone but it works for me. Because I do not want to spend my time fooling with hardware I buy off the shelf. Its a matter of priorities.
Now money is tight in schools and nowhere is it more tight then in technology. So building your own is doubly tempting. What I did when I was a school technology director was to buy top of the line off the shelf hardware for key applications - the things we absolutely needed to work everyday. The servers were designed and engineered as servers because a number of people who ran servers for businesses recommended that there was a lot of difference there. That worked out great for us. The business manager had a top of the line off the shelf name brand. Getting receipts in, payments (especially payroll) and managing the business of education was key.
In the labs and for many of the others we went down a notch. We didn't build our own (employee time was valuable and we paid students for their work - rightly so I think) but we didn't go with the most expensive brand names. That worked as well. We did maintain a spare parts inventory - mostly disk drives, keyboards, mice and some optical drives - and that let us repair most systems pretty quickly. But we started with professionally engineered systems and I think that was the way to go.
I imagine that in a vo-tech or other school with a computer hardware program building more computers from parts might work out. There are always trade offs though. I wonder how many of them build their own school buildings and school busses with student labor and in the school shops.