Operating Systems

Operating System Functions

• System tools (programs) used to monitor computer performance, debug problems, or maintain parts of the system.
• A set of libraries or functions which programs may use to perform specific tasks especially relating to interfacing with computer system components.

The operating system makes these interfacing functions along with its other functions operate smoothly and these functions are mostly transparent to the user.

• System crashes and instabilities - These can happen due to a software bug typically in the operating system, although computer programs being run on the operating system can make the system more unstable or may even crash the system by themselves. This varies depending on the type of operating system. A system crash is the act of a system freezing and becoming unresponsive which would cause the user to need to reboot.
• Security flaws - Some software errors leave a door open for the system to be broken into by unauthorized intruders. As these flaws are discovered, unauthorized intruders may try to use these to gain illegal access to your system. Patching these flaws often will help keep your computer system secure. How this is done will be explained later.
• Sometimes errors in the operating system will cause the computer not to work correctly with some peripheral devices such as printers.

• Windows XP Professional Edition - A version used by many businesses on workstations. It has the ability to become a member of a corporate domain.
• Windows XP Home Edition - A lower cost version of Windows XP which is for home use only and should not be used at a business.
• Windows 2000 - A better version of the Windows NT operating system which works well both at home and as a workstation at a business. It includes technologies which allow hardware to be automatically detected and other enhancements over Windows NT.
• Windows ME - A upgraded version from windows 98 but it has been historically plagued with programming errors which may be frustrating for home users.
• Windows 98 - This was produced in two main versions. The first Windows 98 version was plagued with programming errors but the Windows 98 Second Edition which came out later was much better with many errors resolved.
• Windows NT - A version of Windows made specifically for businesses offering better control over workstation capabilities to help network administrators.
• Windows 95 - The first version of Windows after the older Windows 3.x versions offering a better interface and better library functions for programs.

There are other worthwhile types of operating systems not made by Microsoft. The greatest problem with these operating systems lies in the fact that not as many application programs are written for them. However if you can get the type of application programs you are looking for, one of the systems listed below may be a good choice.

• Unix - A system that has been around for many years and it is very stable. It is primary used to be a server rather than a workstation and should not be used by anyone who does not understand the system. It can be difficult to learn. Unix must normally run an a computer made by the same company that produces the software.
• Linux - Linux is similar to Unix in operation but it is free. It also should not be used by anyone who does not understand the system and can be difficult to learn.
• Apple MacIntosh - Most recent versions are based on Unix but it has a good graphical interface so it is both stable (does not crash often or have as many software problems as other systems may have) and easy to learn. One drawback to this system is that it can only be run on Apple produced hardware.

A computer consists of a collection of functions. The functions are things like memory management, input/output processing, computing (arithmetic), making decisions (logic) and tons of others. Each one of these functions has their own "controller", which can be a simple chip, or an entire processor (a collection of chips). So...when you hit a key on the keyboard, the processor on the keyboard processes the key, determining when you have hit the key down, when the key is up, whether the key is being held down etc., and finally assigning a value to the what the key is doing (for example, the value might be "1" if the "a" key is hit, "2" if it is being held down, and "3" for returned back up, the "b" key might be values of "4" for down, "5" for held down and "6" for back up). The first question you might have is "who determines the values.". Well, back in the "old" days, the manufacturer would determine these values - so IBM had one set, DEC (Digital Equipment) had another set, Honeywell had another etc. As the computer environment matured, standards were set up, for instance using ASCII (American Standard Code for Information Interchange) code for things like keyboards, storage and internally in a personal computer. On the mainframe, the code used was/is EBCDIC, (Extended Binary Coded Decimal Interchange). Once these standards were set up, then each piece of the computer (keyboard, memory, logic, storage etc.) could be programmed using the same "code" to communicate with each other. At that point, a centralized approach could be taken to control the communication with each component. This resulted in processors that have "chip sets" that control the communication between the central processor and everything else attached to it. This was great, because now you could buy a computer out of the box that have a keyboard, storage (hard or floppy disk), monitor and other functions that interface together to run. However...most of these computers still had low-level (what we call "machine code") programming so that each of the components had to be programmed separately and then wired together to operate together. They also had to be made by the same company, since that company decided how each one worked. The first operating systems were made by programmers who got tired of doing everything at a low level. They started writing programs that would do some analysis of the environment, and then use that analysis. For instance, you might determine does the computer I am running on have a hard disk? if the answer is yes, set a flag in the processor that says hard disk is attached. When it came time to do something with the hard disk, you would issue a command that said "is the hard disk powered up and available?", if the answer was yes, you could then issue a command , like read the disk (and the read would be to a very specific address on the hard disk), then another command that said "was the read successful?", if the answer was yes, then whatever was read would be copied to internal memory. Then another "read" would be issued to get the rest of the information you are looking for - since each "read" would only read 1 byte of information. As you can guess, you had to know the length of the data that you wanted to read, so you could loop through the read cycle for that number of bytes.

Now doing something like a read cycle has a lot of steps to it (is there disk? is it ready? did I read? was it good? am I done reading?), so programmers first invented loops that did that function. Now these low level functions are handled not from the central processor, but in each device (keyboard, disk etc.). The "loops" the programmers used were the basis for todays "device drivers." Device drivers provide the interface between the central processor and the "peripherals" of the computer like keyboard, disk, monitor etc. Device drivers take into account vendor peculiarities (they call them "extensions) that vendors use on their device, and allows the central processor to communicate with any peripheral at all, as long as there is a device driver that recognizes that device, what standard it is using and what vendor differences there are from the standard.

Now...what does all the previous stuff have to do with operating systems? An operating system consists of two main parts, one that controls the computer/peripherals and another that controls the environment, which we commonly call a "GUI", a Graphic User Interface (we recognize this as "windows" or "Linux").

So...what is the requirements of an operating system? The requirements are that you have a computer with a central processing unit and peripherals (and yes, networking is a "peripheral" - we use network cards or chips). You need to know how each internal component relates to each other (usually on a "bus", but some have their own means of hooking up), and how each peripheral works. An operating system starts with a "boot." When you power on the computer, there is nothing known about that computer - the memory is empty, the internal environment is unknown, the peripherals are unknown. So, the first thing - operating systems are written to handle a particular environment - in the case of Windows and Linux, they usually handle Intel-based environments, but can be written to handle other environments (such as mainframe processors, or non-intel environments). The "boot" identifies the environment, internal components and peripherals, and sets up an inventory of functions that it can perform (we call this inventory "vectors" and set up a vector table for each computer). This inventory includes what is available, is it ready?, how to talk to it (what device driver to use) and any special information that function needs. After doing the inventory, and testing most components, it brings up the GUI, allowing the user to do what the GUI allows (execute programs/packages, control the computer etc.). The computer can also be controlled at the "command line" level, with or without the GUI. In either case (GUI or command line), you are communicating with the same operating system.