The announcement by Google that it will be offering an operating system has stirred lots of debate. But what makes an operating system good or bad? What are the tradeoffs? Here are 10 measures on the operating system yardstick.

1. Security. This can be a show stopper right at the beginning. Can exploits be identified and contained? Can incoming data be confined to a segment where the errant data cannot reach other parts of the operating system and applications? Can a user set degrees of control and encryption of the data on their computer? How are updates received, authenticated and installed? Lots of questions here, but security is step one.

2. Does it work? Simple question but a viable question. What happens when you install the operating system? Can it run in its own disk area without messing around with other installed systems? Does the installation keep you informed all the way of what is going on with the installation?

3. Is it compatible with the past? If you have to toss out your computer to run the new system, then you are buying a marketing product rather than a technology product. How about an operating system that might run slower on an older system but still runs? If you really want to run your operating system faster, then you can decide to buy a new computer.

4. Is it part of a family? Mobility, cloud computing, notebooks and servers all need to play together. Maybe the user doesn't care too much about the underlying operating system, but the IT manager cares a lot about training, system management and future developments. The operating system, applications and hardware requirements all need a road map.

5. Application development. In business it is all about the applications that help a business complete its business strategy. Operating systems and application environments are intimately connected. Building an application development team is no simple matter and having to retrain that team for a new operating system environment is simply too big a task for many businesses to undertake.

6. Cost, not price. Business and technology managers care about the lifetime cost of their technology products. A low or free initial price is not much good if the product deadends or the vendor disappears. What is the 10-year horizon and 10-year sunset of the operating system? Will current applications run on the new system? What about hardware requirements? What about future developments? These factors all play in the cost equation. License and liability costs are also part of this consideration.

7. Future trends. Five years ago not many people were actively installing virtualized IT infrastructures. Now it is on nearly everyone's radar. While you cannot always predict the future, you can take educated guesses from current trends. How does a new operating system fit into those trends?

8. Help. What happens when you run into unforeseen problems? Every vendor is your best friend before a sale, but the real test is how helpful they are when a problem arises. This consideration is part of the operating system equation as the vendor may be your only lifeline when things head south.

9. Stability. This isn't the financial stability of the vendor, this is the stability of the product. What resources are available for bug fixes, crash reporting and application incompatibilities? The rule in the past was never buy Version 1.0 of anything and instead wait a year for all the bugs to be squashed. Vendors hate this rule.

10. The environment. No, not that environment we should all be concerned about. This is the environment where the operating system will function. Can you introduce the new operating system in the IT environment gradually without disrupting the other pieces of the network?