Operating System Evaluation: The Software Solution

THE SOFTWARE SOLUTION:

We’ve already determined that we have a problem. Gasworks currently operates its daily functions, as well as large scale projects, on a Windows 3.1 platform. The problems associated with this operating system, which have lead us onto the path of change, were outlined earlier. It is now our decision to take a look as to what is currently out there, and make an informed decision as to which operating system should control the hardware of the Gasworks. We will be looking at a number of criteria before we come to our final decision. Functionality, speed, user training, hardware requirements, and transference of existing software are just a few of the areas that require investigation. Once all of the pros and cons of each operating system are examined, only then can we make the best decision about which operating system would be best suited for Gasworks for a successful future.

We will examine such systems as Windows ’95, DOS 6.x, NEXTSTEP, and even toy with the idea of maintaining our existing Windows 3.1 system. Due to the length and duration of time allotted for this project, we can only consider these four operating systems. However, if more time and resources were to be invested into this project, our emphasis could include many more systems. Unfortunately, time and resources are a factor, and only these four can be examined. These four selections represent our top four choices in operating systems which we estimated to be the best candidates for assuming the role of Gasworks’ future platform. However, since the scope of this project is limited, our choice cannot be regarded as final and correct. We do feel, however, that our data collected will be a satisfactory means for us to reach a well-educated decision.

OVERVIEW OF SYSTEMS:

Let us first begin with the Disk Operating System (DOS) version 6.x. DOS was extremely popular during the mid and late eighties, before the explosion of the graphical user interface (GUI). It used command line instructions to execute various programs and procedures. For example, the user would be faced with a prompt, or a command line input point, that he or she could use to indicate which program was to be run. If the user wanted to format a floppy disk in a disk drive named "a:", they would type in the command: "format a:". This type of a system relies on learned commands to maneuver throughout the DOS file structure. This could be very detrimental to our employees at Gasworks who are used to the GUI environment and are not accustomed to entering commands, but using visual devices to run programs and manipulate files.

Many software applications were developed for this platform, but newer versions lead away from the DOS environment and into a Windows or GUI environment. There are word processors, spreadsheets, graphical design packages, accounting programs, and countless other applications readily available to the DOS user.

Our current system is Windows 3.1. This version of the heavily popular Microsoft operating system was an industry standard for a number of years. It brought the GUI to millions of people worldwide who, until then, were used to the command based system of DOS. Windows used graphical icons or pictures to represent various programs and system tasks. A new user could easily associate with the different icons and become more familiar with the system, thus lowering the learning curve. The use of commands like in the DOS environment were still available, but remained hidden behind the glamorous user interface. Due to the overwhelming popularity of Windows 3.1, a plethora of software was created, or in many cases, upgraded to meet the new graphical needs. For example, WordPerfect for DOS became WordPerfect for Windows, and in many cases, leaving the DOS counterpart behind in the cloud of the Windows popularity. Windows 3.1 is still widely used due to the fact that many users are not yet comfortable to switch to ’95, and also because of the hardware constraints that Windows ’95 demands. Windows 3.1 uses a 16-bit architecture for its operation, however towards the release of Windows ’95, version 3.1 was available in a 32-bit version.

Windows ’95 is a new operating system introduced to the world this past summer. With all of the media hype and the juggernaut company Microsoft backing it, Windows ’95 is bound to be the most used operating system of the near future. It uses a 32-bit architecture for its operation, and with the development of more and more 32-bit applications and faster computers to run them, Windows ’95 is a very fast and powerful operating system. It uses many of the concepts employed by Windows 3.1, but adds a level of consistency where 3.1 was lacking. Shortcuts, icons or links to programs, can be used throughout the system to give the user speedier access to their applications. One great advantage of Windows ’95 is its plug-and-play capability. If new hardware is to be added to a system running Windows ’95, the system almost instantly recognizes the new addition and loads drivers necessary for it to run. It’s as easy as that. In fact the whole operating system acts in a similar fashion. It is easy enough to use so that a novice user can feel comfortable manipulating the interface, and at the same time allows the power-user to perform sophisticated and advanced tasks. One drawback of Windows ’95 is its hardware requirements. To run smoothly (and reasonably quick), a Pentium processor and 8 Megabytes of memory are required. This technology may be acceptable to the home PC market, but may be restrictive to a company with a vast diversity of legacy machines not quite up to the level that Windows ’95 has set for itself. As the technology, the accessibility to the technology, and the prices reach an acceptable level, Windows ’95 could very well be the most used operating system that the world has seen, even outshining DOS’s late eighties usage and Windows 3.1’s domination of the early nineties.

The last operating system in our consideration is NEXTSTEP, and operating system developed by the NeXT company. NEXTSTEP is an operating system based on the Berkeley standard UNIX operating system. It is analogous to the way Windows relates to DOS. One is a GUI based on the command line structure of the other. NEXTSTEP offers a unique object-orientated core whereby all of the subsections of the operating system are similar modules or objects. This approach to graphical user design helps to maintain interface consistency, which can aid the novice user to learn the system more readily. Unfortunately this point may be moot because of NEXTSTEP’s use of the UNIX operating system. For a lot of users, this system may as well be from Mars. Most users today have a fair grasp of DOS, a very strong understanding of Microsoft Windows systems (whether it be the 3.1 or ’95 version), most have no understanding at all of the UNIX operating system. NEXTSTEP has very strong client/server capabilities, able to interact with DOS and Macintosh file systems. It may even be better than today’s answer to client/server applications on a Windows system. Novell NetWare is widely used to network today’s PC’s running the Windows platform. The key advantage that NetWare holds over NEXTSTEP is the fact that it utilizes the MS-DOS/Windows platform. Few applications are readily available in the NEXTSTEP environment as compared to the Windows system. The learning curve with NEXTSTEP would be undoubtedly high, and the price of scarce software and hardware are factors harming the choice of NEXTSTEP. As good as the interface may be, it is still not as readily available and supported as the Windows platform.

BENCH MARKS (SPEED)

Speed is a consideration when examining a system. It has to be able to run large and robust programs with very little delay to the impatient user. The MS-DOS system is a slow operating system when comparing it to the others. Slow in the sense that to run the same graphical programs that a GUI does, DOS would take a much greater time. DOS is fast for its command line operations since no graphical overhead is needed, and DOS is not a mulit-tasking environment so it can concentrate on one program at a time. The bonus of using a command line system is that the programs that it runs are not usually based in a graphical user interface, and speed can sometimes overlooked. Unfortunately, Gasworks is looking for an operating system for the future and would rather not sacrifice speed. As programs become larger and more intricate, speed will play a large role in the success of the system.

Windows 3.1 is a system with a adequate computing power. Its speed is hindered by its 16-bit processing. Mostly all of today’s hardware is based on 32-bit architecture and a 16-bit operating system would just slow down the performance of the hardware. On the positive side, it contains good memory management facilities which speed up the whole process. The 32-bit version of this system is a valid improvement, however Gasworks currently has a majority of 16-bit versions.

Windows ’95 is probably the best of the bunch when it comes to speed. It uses a 32-bit design which maximizes the potential of the hardware. The memory management is superior to that of version 3.1, allowing multiple processes to be run simultaneously (multitasking) in a seamless fashion. Since a prerequisite for using Windows ’95 is a fast machine, the speed factor seems almost negligent when considering Windows ’95 as an operating system.

Having never used a NeXT machine running NEXTSTEP, we here at Gasworks had to rely on documentation to conduct our investigation into this operating system. After reading several sources of information concerning NEXTSTEP, no definite speed values were mentioned which made us a little suspicious. If it was the fastest operating system on the market, they would probably be bragging about it. However with all fairness to NeXT, we would like to continue to investigate the speed of NEXTSTEP to make an educated decision.

PRICE

Probably the most important issue to investigate when deciding on an operating system is the price to roll it out. Not only is the price of the software licenses crucial, but also as important is the cost of support and user training. Also short term and long term prince impacts must also be examined. What may seem cost efficient in the short term may turn out to be a financial nightmare in the future.

DOS is currently installed on a great majority of machines at Gasworks. It was the platform used before Windows was added on top of it. Most users are somewhat familiar with DOS commands, so training may be reduced. The main concern with respect to price lies in the future. With the growing popularity of GUIs on today’s PCs and workstations, less and less DOS based software will be produced, driving up the cost of needed software. Also, Novell NetWare, which is predominantly DOS based, may die out in the future, making way for a Windows-based counterpart. This could dramatically impact the DOS client/server market and directly impact the price of a DOS based operating system.

Perhaps the most cost-effective operating system to plan our future around would be to remain with the existing system. Little to no user training would be needed since the employees have gathered knowledge about this operating system over the past few years. Another cost bonus is that fact that mostly all of our software is currently running upon this platform. We have support that are familiar with the Window 3.1 environment, with vast array of invaluable knowledge about its strong and weak points. One main concern regarding price, is the limited software of the future. Big brother has arrived in the form of Windows ’95, and most developers will undoubtedly be focusing their attention to this platform in the near future. Soon, Windows 3.1 may fall to the wayside, just as it predecessors have (Windows 3.0, 2.x, etc.). Although Windows 3.1 may be the most cost effective platform to work with in the near future, the distant future may show that the road that 3.1 has paved for itself leads straight into a money pit.

Windows ’95 seems to be the talk of the town these days. Some love it, some hate it, some fear it, and others embrace it. Without question, Microsoft is the muscle in today’s computing industry and will use its power to gain even more market share with Windows ’95. With ’95 selling like it is, the supply will soon catch up to the demand for powerful programs to run upon its 32-bit architecture, driving their prices to a competitive level. An advantage of currently using Windows 3.1 is that, if we were to choose ’95, the whole software package need not be purchased. Rather, a cheaper upgrade can be obtained to achieve the same results. In the long run, Windows ’95 seems to be the sensible solution to an operating system platform of the future.

When reviewing NEXTSTEP for a cost analysis, it was very difficult to obtain a bottom line price for the operating system package. In fact, no price could be found at all. The above systems could be purchased for around the $100.00 price point, but from rumors (again, no definite prices were found), NEXTSTEP’s price tag is substantially heftier. Apart from the base cost of the operating system, support will be a large sum of resources since NEXTSTEP is a specialized system and low supply and high demand equals expensive support. Also the cost associated with converting our employee base who is firmly grounded in the Windows environment into a UNIX based GUI may be the largest price yet. From screen snapshots of the operating system, the interface looks very sophisticated and alien to longtime Windows users. It is the high cost of support and training that may cause NEXTSTEP to be overlooked.

Information obtained from a reliable source states that by the year 2000, Microsoft will control 83% of the computing industry market share, UNIX at 8%, and all other companies battling for the remaining 7% (even including the fallen giant IBM). With these bold predictions, it would seem cost effective to follow the industry standard rather than pursue a platform with minimal user impact.

FUNCTIONALITY

MS-DOS has all of the functionality that will be required by Gasworks including the usual word processors and spreadsheets, as well as client/server software allowing for extensive networking capabilities. However it seems that DOS is on the way out and Windows is here to stay. With this fact of life upon us it remain to be seen if the functionality for DOS will increase or decrease. With the release of Windows ’95, DOS is no longer needed (as it was in previous versions) and may diminish with the growing popularity of Windows ’95.

Windows 3.1 scores very high on the functionality scale in that it is very robust and the applications available for it cover a wide scope of all software available. As mentioned above, it is unlikely that 3.1 applications will continue to be generated due to the fact that Windows ’95 offers a quicker and more efficient way of handling its information, and as the demand for speed grows, the desire to use Windows 3.1 (and for that matter develop software to run on that platform or to upgrade existing software) will wither away.

Windows ’95 offers the best functionality of all systems tested. Its plug-and-play allows for virtually and hardware to be added or removed with ease, and the beauty is that the operating system will actually help you in this process whereas in previous versions, one had to struggle to get any hardware added or deleted. Windows ’95 offers all of the function available to Windows 3.1, but with all of the hype and enthusiasm, its functionality will grow as more and more developers choose ’95 as a suitable (and profitable) solution to their needs.

NEXTSTEP offered an operating system with a vast scope of functionality. It is strongly object-orientated and offers these modules to developers to organize software in a consistent manner. It is also very proficient in client/server applications, arguably more so than DOS or Windows systems. However, the interest in a UNIX based system is not such to expand its functionality to the degree that Windows ’95 is.

LEARNING CURVE

In the past, an employee who was hired at Gasworks was subject to several days of training concentrating on the tools that most used to complete his or her tasks at work. This would have included extensive Windows 3.1 training. If we chose to stay with Windows 3.1, we would need no turnaround time for learning since our employees are presently adequately trained.

Windows ’95 would present us with a situation where training was needed, but not on a grand scale. Most users find Windows ’95 much easier to use because of its interface consistency, response time, shortcuts, etc.

MS-DOS would be a little more difficult to train our employees to use. It would seem like a step backwards, going from a GUI to a command based system. Many users who have not had previous experience with DOS may be hesitant to learn it, but most gradually begin to remember its many commands and switches.

NEXTSTEP would present the greatest problem with training our employees and their subsequent learning of the system. In ways it is very similar to Windows ’95. Some of the graphical widgets look familiar and there may be some positive transference of knowledge to NEXTSTEP, however, being based in UNIX, will require extensive training. DOS and Windows are closely linked and most employees feel somewhat comfortable working between the two systems. There is a huge gap between DOS/Windows knowledge and UNIX/NEXTSTEP knowledge. The curve would represent a poor ability to learn a foreign system.

TESTING PROCESS AND RESULTS

The above chart displays the qualitative results gathered from the comparison of the four operating systems in question. The numbers are based on a 5-point scale where 5 is a favorable result and 1 is a poor result.

The chart clearly shows Windows ’95 to be the leading choice in operating systems. It scored perfect in functionality and speed, and above average in price and learning curve. After weighing the pros and cons of each operating system, ’95 shows its dominance when compared to the others. NEXTSTEP came in a disappointing last place. This is not necessarily due to the quality of the interface and the programs that it provides, rather its poor showing is a direct reflecting of public interest in the operating system. With scarce support facilities and uncertain future functionality, it would make sense to shop elsewhere. Windows ’95 had stiff competition from its forefather version 3.1. It seems almost justifiable to leave well enough alone, however the benefits of ’95 could not be ignored. DOS, even though it placed higher than NEXTSTEP, could not compare to the strong Microsoft GUIs. By using this qualitative data, we here at Gasworks have been placed into the situation whereby we now can make an educated decision as to which operating system will lead us into the future. Windows ’95 is the golden road to a prosperous future, and Gasworks will decide to enter onto that road to share in the success.

Back to Table of Contents