Idea behind thin clients

You have the idea about thin clients all wrong. The point isn't to sink one-time hardware costs for the client, but the TCO of the whole solution, mainly:

Low hardware maintenance costs. No movable parts and very quiet. If it breaks down, you just replace it without any (re)installation.

Low software maintenance costs. Instead of maintaining hundreds of machines, you just maintain a couple of servers.

Consistency. A user can use ANY of the machines just like his own.

Centralisation. Well, this is the main point from which all of the above can be derived. You can also centralize backups, internet connectivity, application testing, etc.

Why aren't the thin clients used more you ask? This was answered in one of the previous posts: Windows thin client solutions suck and the server costs are horribly expensive. And people want Windows.

If you strip the Microsoft requirement, there shouldn't be any problem. In fact I am writing this post from my own thin client, which I have been using for almost a year. I bought the thingy (Igel-J thin client) on Ebay for like 220�, and run Red Hat Linux 9 on the terminal server (which cost under 700� and is in a climatized server room). Add a high quality TFT monitor and you get an el-cheapo el-coolo solution. The only thing that I can't do on it is to watch movies, but I have a separate specialized machine for that.

Instead of thin clients, I would like to propose a different fad: a Windows solution that actually works.