Database Systems Concepts - Essay Example

An operating system should aim for two things: It should manage input and output devices, and it should control, supervise and empower (which each have separate sub-goals and are sometimes in conflict) users as well as administrators and people connected on the network to use the hardware to run software. Efficiency helps these goals because it reduces lag in input and output and because it reduces the overhead of the OS thus allowing people to run more programs and run them with less slowdown and issues.

However, it is unquestionable that DOS would be far more efficient than Windows 7. Assume for a moment that a DOS-like text-only input with no or few visual input systems could run every element of modern software. Most consumers would still prefer not to spend their time mucking about in command prompts unless it's absolutely necessary. Similarly, Linux users often get better machine-specific functionality by compiling their own kernel, but for most users, “good enough” is better because an OS like Windows designed to run on the lowest common denominator takes more of their time.

Very few people like to struggle with computers: They want to quickly install the OS and get to work on what they're doing. Their time is valuable either to them or to their organization, and so it's actually inefficient to waste their time with OS “efficiency”. User friendliness is one of the keys to the success of Macintosh. Many experienced PS users smart at Macs because they feel that they are being too constrained, but they are often lifetime computer users. Many people don't have the temperament or interest.

“Efficiency” in the form of less graphical displays and overhead is wasteful for these customers because that overhead gives them a better, more pleasurable and more efficient experience. Power consumption is another area where OS overhead could be acceptable. Controlling for power does cost computational time, but it also saves customers money. On laptops in particular, having power management tools, which are an additional process running and thus introduce inefficiencies and bloat as well as risks of system failure in terms of power consumption software errors, are worth it because they allow users more time with their laptops.

Allocational tools can also introduce temporary overhead, but if customers are intending to run dozens of programs, as most are, having smart allocation is an investment that pays off. In fact, many OS tools are basically investments with short-term inefficiencies that lead to long-term computational efficiency. Drivers with video cards, for example, clearly cost some degree of computational time and lead to bloat (and anyone who has struggled with failing drivers or Direct X compatibility issues knows that errors in these arenas can be catastrophic), but they allow the usage of additional RAM devoted to graphical elements which is necessary to run most modern games or modeling software.

One of the more controversial elements of system overhead is security. In essence, many OS designs don't trust their end user to be safe. Firewalls with highly restrictive settings, anti-viruses that are hard to get rid of, built in anti-spyware, permissions management systems. these improve security, with some risk of exploitation (since every security measure can also be a security loophole), for the lowest common denominator user with little skill