Making the Move to UNIXLook before you leap |
|
From "Migration", Access to Wang, August 1993 |
|
[ Prior Article ] [ Return to the Catalog of articles ] [ Next Article ] |
Judging from the results of recent surveys, quite of few of you are considering a move to the Unix environment from your present position on the VS. Unix systems present a real opportunity for performance improvement and a cost-effective mini-based substitute for the VS, and provide the most likely landing spot for those considering migration - the orderly movement of systems to another platform. Most other conversion choices - including LAN-based PCs and other proprietary minis - either offer no tangible improvement in performance or require a large conversion effort to implement. And if you plan carefully, the movement from VS to new Unix system can be a smooth transition - possibly blending the best of both worlds.
If this describes your situation, welcome to the world of VS/Unix coexistence and migration. This column is dedicated to your needs. Over time I hope to help describe the options available for coexistence, assist in understanding the Unix environment, and share experiences as more VS commercial applications hit the Unix world. The information will be presented from the perspective of a VS user - a critical user, at that.
In this debut column, I will review some of the changes that can be expected in the Unix world and how you can begin to prepare for them. I will also discuss some of the reasons that management is more involved in this system decision than before. Subsequent articles will cover the technical differences between the environments and some of the choices that must be made.
Talk to the managers of VS shops about a future switch to any other environment and you're likely to hear these statements:
The change was forced on us from top management; we didn't have a say in the decision.
Why can't we just keep the VS?
By now, the answer to the second question may seem apparent, as even Wang Labs recommends that customers consider moving from its flagship processor to other systems. The visible financial difficulties of Wang Labs have made it more difficult to defend an otherwise capable environment from any change at all. The big change for many of us is the direct involvement of top management in the decision process; sometimes the decision has been made before the technical merits of the choices are discussed.
If you manage a VS shop that has been in place for some time, it's likely you have seen more difficulty in getting management approval for VS equipment and software in the last few years. Viewed objectively, most VS shops could be said to be very successful - if success is measured in reliable operation and conformance to the requirements of the past. But information systems have grown in importance - or at least in prominence in the financial press - and management readers have come to believe some of the hype, including:
'Proprietary' is a synonym for 'bad'. No one should be locked into any operating system any more.
Your competitors are getting smarter about their use of information. You must keep up. Information should be in the hands of all users that need it, and Open Systems and relational databases are the only way to make that happen.
Naturally, there is some truth to all of these points. The existence of proprietary systems has made transitions between vendors a painful, expensive exercise, and all of us have spent far too much money in the past because of vendor barriers. Remember $4500 terminals? How about all of those buildings that were wired with dual-coax cables for Wang terminals?
There is also substantial evidence that information has become a strategic resource - and not just for companies in the Fortune 1000. Better information access and control can result in more personalized customer contact, more accurate marketing, and other means of increasing effectiveness or reducing costs. In some industries, the ability to communicate electronically with other systems is already considered a requirement for doing business.
The last point of these points reflects an attitude that could be termed the third information revolution. The first revolution occurred in the late 1970's with the introduction of minicomputers like the VS; finally, it was possible to perform real data processing work with a limited technical staff and some end-user control. The second revolution occurred during the proliferation of the PC in the mid-1980's, when information became a tool of managers and other end-users. Finally, these PC users were not satisfied re-keying critical information from print-outs or text files: they wanted the real data, right now. This third revolution demands integration of enterprise data into the familiar tools of the desktop world.
It's certainly a shock for those of us who have paid careful attention to security and the integrity of data to open that data up for anyone to use. The challenge of the new information age is to make certain that the availability of information does not extend past a predetermined organizational circle.
Many data processing professionals view the change to Unix with distrust - fear, even. Some common reasons for this malaise include:
Unix systems require more complex system administration and are known to be security risks.
All of our end users will be forced to use a command prompt.
It's all new to me; I don't want to learn a new environment.
To make this situation perfectly clear, Unix systems are more complex and difficult to maintain, and not just because they're new to all of us. The evolution of Unix from think-tank to academia to commercial applications has left it a large, seemingly untamed collection of cryptic commands and folklore. There is also some reality in the claims of security problems, though most problems appear to be caused by mis-application of the very robust security features or inattention to the detail required by such an environment. Viewed through the same viewpoint, most VS environments are insecure, with most users able to use the DISPLAY utility to view the contents of any production file they have access to and CONTROL files there to help describe the elements of those files. As for the 'look and feel' of Unix, you can take your pick of just about any environment you wish. Most migration approaches will attempt to retain the menu-driven applications familiar to end-users, but the technical environment can be a command prompt, a Graphical User Interface (X terminal), a menu, or some combination. True, it will be necessary for most technical users to understand Unix from the command prompt level, but others may dwell where they please.
In the need to learn a new environment, no VS user goes into a command prompt environment willingly. Unix is undoubtedly one of the richest environments available, and its basic form can be extended and modified to fit many other expectations. There's power in those command-line hieroglyphics - power you may not believe until you master a few. I recommend you enter into this process with as close to an open mind as is possible.
As you may have gathered, the transition to Unix will require some work and a good deal of time. You will probably find some formal training helpful; community colleges are a good choice for this. Book publishers are pumping out Unix texts of all sorts. Unix magazines will offer current information - though be prepared to suffer through the strange vocabulary for a while. Some Unix user groups are helpful, but most are composed entirely of the socially challenged.
But the best training is practice. It will not be possible to even get passing familiarity with Unix usage without contact with the environment. If you do not yet have a Unix system, consider purchasing one of the several available Unix-like environments for PCs as a training ground.
I hope you find (as I have) that the power and flexibility of Unix can awaken some of that career interest that might have been caged too long.
Copyright © 1993 Dennis S. Barnes
Reprints of this article are permitted without notification
if the source of the information is clearly identified