Operating System Evaluation: Task Description


Virtually everyone in the company will be effected by the decision of which operating system to implement. The user base within the company has a wide spectrum of computer knowledge, from illiterate to super guru, and the desktop platforms are used for a variety of tasks. Some of these include common applications such as word processing, email, and spreadsheets, to the relatively obscure, such as specialized cost estimating, CAD, and various in-house packages. Thus, the operating system must be extremely flexible and adaptable, easy to use, yet offer powerful and advanced features for those who require it.

With 3000 employees currently using Windows 3.1 on their desktop machines, it is imperative that these employees be able to efficiently use the new operating system with little training (as with this quantity of employees, training would prove to be a massive expense in money and man hours). Additionally, an employee who cannot use the tools provided efficiently costs the company money. Factors considered under user ability include similarity to Windows 3.1 (as this is what they are familiar with), as well as outside exposure. For example, it is likely that numerous employees will be upgrading to Windows 95 on their home computers, thus reducing or even eliminating the learning curve in the office environment.

Currently a large base of MS Windows 3.1 and Windows applications knowledge is present in The Gasworks to support our desktop environment. In order to best make use of this existing knowledge, it is imperative that the new operating system function similarly to Windows 3.1, and that the products currently in use with this environment are available on the new operating system. Retraining of support personnel would be extremely costly, and would likely necessitate the bringing in of outside contractors until our own people can be brought up to speed and are experienced enough to satisfy our needs.

After the large learning and training issues, legacy system compatibility becomes a prominent factor. Issues such as application changes, application availability, and application compatibility become paramount. Application changes entail converting application licenses from one platform to another; including upgrade fees and licensing arrangements. However, first the question of whether or not applications are even available for the new platform must be considered. If not, then how can legacy data be accessed? Must users be retrained on new applications? Clearly after many years of using Windows 3.1 and applications in that environment, there is a massive store of data native to those applications. This data must remain accessible under the new operating system and its applications.

Existing hardware must also be taken into consideration. The majority of desktop machines within the company are 486/50 and 486/66’s. If an operating system cannot run at all or at a satisfactory speed on this architecture, upgrading or replacing these machines to meet the required operating system specifications would constitute a great expense. Additionally, most machines are currently equipped with 8M of RAM; while most of the newer leading edge operating systems require more than this to operate comfortably, how much more becomes an important issue. A great deal of the peripherals currently in use around The Gasworks conform to the SCSI standard. It is crucial that the new operating system be capable of taking advantage of these devices. Similarly, the availability of device drivers and support for other common hardware components is strongly desired.

Installation cannot be overlooked in O/S selection. For example, an operating system that can be installed on a server and automatically pulled down by client machines would be a HUGE advantage over physically having to visit every one of the approximately 3000 desktop machines across the province. However, as this is not likely regardless of the operating system, it is important that the installation process be as seamless as possible, and, ideally would upgrade from the existing Windows 3.1 platform using as close as possible whatever the current configurations may be on each machine.

The features and capabilities of the operating system from a user’s point of view is also important. For example, does it support data synchronization between desktop machine and laptop machine? Does it allow for plug and play? Does it provide an intuitive interface and extensive help? Are extensive multitasking capabilities incorporated? Is the operating system robust and fault tolerant? Can users easily connect to company networks? These factors all have a great influence on employee productivity, as well as have a large impact on the demands of end user support.

Administrative features are also substantially weighted. Point in case, an operating system providing facilities for remote support would be beneficial to our support staff. The ability to grant access to various applications from one central server, security capabilities, and auto configuration of hardware devices are all important. Under security requirements, it is desired that the operating system have features such as secure logins, login controls, network password protection (i.e. encrypted passwords being sent across the network), C2 and B2 security standard compliance, password aging, and robust, assignable security levels to various files. While some of these features may not be in the domain of some operating systems, it is felt that networkability is a major feature that The Gasworks is looking for; the current Novell NOS is seen to have a very limited future within the company and will in fact likely be phased out over the next few years.

Most of the in-house applications currently in use within the organization have been created with either Visual Basic, C++, Powerbuilder, and Oracle. In order to avoid the costs of porting these applications over to other environments, it is hoped that the new operating system will have compilers available for these development languages. If they are not supported by the new operating system, then the costs associated with re-training or bringing in external assistance to use different tools must be incurred.

Two specific feature areas being evaluated are networking capabilities and file system management. It is hoped that the new operating system will have internal support for TCP/IP, Ethernet and Token Ring environments, specifically supporting the network interface cards currently used within the company. Other protocols, such as IPX/SPX, NetBEUI, AppleTalk, and Pathworks are desirable and provide flexibility but are less important. As part of the networking component, integrated internet applications, such as FTP, News, Mail, etc., would be an added benefit, as would PPP and SLIP protocols for easy internet access.

File system management involves issues such as long file names, maximum secondary storage capacity, and maximum partition sizes. Additional considerations include the ability to store files and/or volumes across multiple drives, backup capabilities, and, if the operating system has server capabilities, RAID technology. RAID refers to Redundant Array of Independent Disks, and provides excellent fault tolerance across systems with a minimum of three hard disks at a cost of slightly reduced user storage space. However there are many third party products available to implement this technology, and as such it is not critical to be included as part of the operating system.

Compatibility with other operating systems would be beneficial. Allowing for a wider range of applications, and much more data portability, it will also allow for the flexibility to more easily update operating systems in the future. However, one danger that must be avoided is an operating system primarily being used to emulate a different operating system. There is little point in doing so; the "middle man" can be eliminated with little repercussion. An example of this would be OS/2; if selected as the operating system of choice, there would be little incentive to take advantage of it’s native operating system advantages since all legacy applications will execute perfectly in OS/2’s Windows 3.1 mode.

The ability of the operating system to run on multiple hardware architectures would be extremely beneficial. This would allow identical software applications to execute regardless of the hardware, permitting great advances in standardization across the company, and all the benefits associated with that prospect (reduced licensing costs, more universal support, increased data portability, etc.).

An operating system’s future is also a large factor under consideration. If an operating system is chosen, and it consequently ceases production, there will be little support and application development available, in effect necessitating another operating system change. Following the crowd boasts the advantages of having a wide range of available software, a wide range of support for that software, and a wealth of available expertise to drawn upon.

Related to some of the previously mentioned factors is support. An operating system that is widely familiar will be easier to find support for, both internal and external to the company. Support from the vendor is also of great importance; clearly a supplier with an effective and inexpensive support policy will be favored over those who are reputed to have poor or costly technical assistance.

The last, and most obvious factor involved is the outright cost of the operating system software. However this is not quite as straight forward as comparing numbers, as features, configurations, and included applications vary GREATLY between operating systems, and even within different marketed versions of the same operating system. As The Gasworks is a very profitable organization, purchasing a more costly operating system is not a problem, even if it may exceed the requirements for most end users.


Back to Table of Contents