The Software Commons Revisited

It turns out that The Software Commons is more compelling than I had first thought.

The Software Commons Revisited

When I wrote The Software Commons article, I was focusing on Open Source on the desktop computers in offices. However, I had been looking at it as a direct replacement for Microsoft Windows and Office. Some recent articles have made me realize that the financial advantages of using Linux are even larger than I had been thinking.

Years ago, many companies used mainframes for their computing power. Everyone had a dumb terminal on their desk to act as a keyboard and monitor for the mainframe. This model is convenient, because it provided easy administration. The IT folks just had to keep that one big machine running well, and everyone’s software was always in sync. The original terminals were just plain text based and the user interfaces were not always that easy to work with. Unix and the X window system added a graphical interface to the client/server paradigm, but X based systems were not very easy to use either.
Now, we come to today. Many office workers have computers on their desks that are more powerful than the mainframes and Unix workstations of a few years back. The applications that most people run at the office hardly even touch the power of the computer they’re using. In fact, there’s probably enough CPU horsepower there that a few people could share that one computer if they had a way to do so.
And, that’s precisely what is happening at some offices now. A single fairly powerful PC can comfortably handle quite a few users. In fact, there’s a city government in Florida that has 200 people using one (albeit quite powerful) PC simultaneously. Each user can have a slow PC on their desk, because all it is doing is acting as an X terminal. One writer talked about adding a new user to their network for under $30, not including keyboard, mouse and monitor. The graphical environments and applications on Linux and X today are fairly similar to what you find on Windows, so the standalone PC model doesn’t have the ease of use advantages that it once did.
So, a typical office worker can do what they need to do, and do it productively, on much cheaper hardware and using free software. Additionally, the cost of administration is lower because only a small number of servers need to be managed (and backed up). And, as an added bonus, anyone can conveniently work from anywhere in the office.
This, of course, is not some brand new model. People in the mainframe and the Unix worlds have been doing this for years. The difference is that the user interfaces and applications available for this platform are as sophisticated, productive and easy to use as the ones for Windows.
As I mentioned in The Software Commons, all office workers need a set of baseline software to work with: operating system, word processor, spreadsheet, presentations. Upgrades to Microsoft Office are less compelling now, because the earlier versions of the product handle nearly every task that people commonly need to do. Each upgrade to Windows and Office seem to require more computing resources, as well. So, the cost of upgrading is often larger than just the cost of upgrade licenses, because there are new hardware and hardware deploymnet costs involved as well. But, if everyone is using X terminals, the PCs on people’s desks can remain unchanged as more sophisticated software is rolled out on the server.
Even a small office can save thousands by adopting a model like this. If a Linux distributor were to package up a distribution that has preconfigured server and desktop software for this type of environment, that could be a very powerful product. It would also tie in well with consulting services to get everything set up, making recommendations for future hardware purchases, and convert data from old formats to the new.