Where will the mainframe be in 10 years? We know from the history >of ISAM
that it will take 30+ years to die after the vendor announces >its demise
and the timer has not started yet.
I'm not sure I understand the question. The concept of a "mainframe" will
certainly be dominant, however that doesn't mean anything with respect to a
particular platform. Systems have always evolved from simple to more
complex implementations, so there is no reason to believe this will change.
If "mainframe" is being used synonymously with "z/OS", then certainly we can
be confident that this will be substantially different. How something like
"z/OS" evolves is clearly an open question, although we can surmise that
there will be a continuing move to speed up data access and so major changes
in data access, archive, etc will continue in a variety of ways. As was
indicated in a previous post, the need to perform large volumes of data
requests will not decrease, so the question is really .... what are the
driving factors in a platform's future?
I believe that several conditions exist:
1. There is a limit to the complexity a personal system can tolerate due to
the lack of expertise in the general user, so this becomes a limiting factor
in deployment. These systems will tend to remain portals into larger
processing environments and personal applications.
2. Mid-range systems and large systems will all continue to grow, so the
same old question exists ... where is the cut-off for transitioning between
the two environments? Companies in this situation will always be changing
since they have more choice in the matter . Once again, it becomes largely
a question of systems management rather than technological capability.
Distribution versus consolidation is an ongoing cycle, so I don't expect any
permanent changes in this debate. Business forces regarding the perception
of saved money or some other reason will continue to drive these choices
whether they are realistic or not.
3. Given enough technical expertise virtually any platform may be a
suitable solution for a given problem, so part of the viability will be the
availability of expertise. While there is a fair amount of technical
expertise regarding systems environments, I believe we are quite deficient
in applications development expertise. This may simply be a lack of
exposure on my part, but I am somewhat disappointed in the lack of knowledge
shown by most applications developers regarding exploitation opportunities
in the various platforms. In my opinion, this will be the primary limiting
factor in platform viability. If applications are not developed, then the
platform dies .... regardless of how good it may be.
In short, the mainframe is here to stay. However, depending on what is
meant by the term, we already know that any existing operating system won't
exist in its present form in 10 years, so ultimately this is a guess
regarding the evolutionary direction of such systems. Regardless of the
platform choices available, migrating applications is not a trivial task
(especially for large organizations), so there is a certain inherent inertia
in making such moves which will also delay radical changes in technology.
It is reasonably safe to conclude that all systems must fulfill basic
systems management objectives, but in the end it will be determined by the
viability and robustness of the applications available on the platform.
Whatever that turns out to be will determine which systems dominate the
landscape in the future.
Adam
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html
