Part 6. Thrashing
Sometimes when you're too busy, you don't get anything done. Have you ever noticed
that? When you start one job, and something more urgent comes along, you drop the
first thing and start the second. Then when something really vital comes up, you
drop the second thing unfinished and move on to the third. Or perhaps you're just
so anxious about completing *something* that you keep dropping whatever you're doing
as soon as the going gets tough, hoping the next thing you try turns out to be much
easier, so at least you'll have finished one thing. As a result, you end up getting
absolutely nothing done!
A computer sometimes gets that way, too, if you overburden it with so many things
to do that the time it takes to change jobs (also known as a context change) keeps
it from finishing any one task. In computerese, this is known as *thrashing*. The
computer keeps attention just long enough to forget the last thing and move on to
the next, but no real work gets done. The computer seems "slow" when in
reality it's moving very, very fast just to stay caught up.
Sometimes a company loses focus and begins doing the same thing, too. If the company
is very large and scattered across many different kinds of products -- such as IBM
-- it may lose focus and not have one consistent message across all fronts. When
things get tough, the response may be similar to that of an overloaded computer.
Each division or department demands attention, and the focus of funding, marketing,
and manpower may begin flitting from project to project in an effort to keep everyone
more or less happy and to make sure no one area is neglected or falls too far behind
in the marketplace.
It seems that IBM has, until recently, had the problem of thrashing to an extreme
degree. First the emphasis was going to be on mainframes, then on PCs, then on operating
systems, then on microprocessors, then on PCs, then on something else. This meant
that if the operating system issue began to be solved, somebody else would feel
left out and need attention. IBM would lose focus just when it was beginning to
succeed, which is what happened in the spring of 1995 when OS/2 was finally beginning
to make some headway. One trade magazine even showed a picture of a blast of air
sweeping into a room, touting the gale-force momentum that was beginning to build
in favor of OS/2.
Just as quickly as that momentum appeared, it somehow evaporated as IBM thrashed
over to something else: electronic commerce. Of course, e-commerce is vital for
IBM's growth and just may become a key ingredient in their long-term success. However,
the danger here is that IBM may at some point begin another cycle of thrashing and
move to something new. The problem with thrashing is that it engenders a certain
lack of credibility and trust in the marketplace. If a software vendor who was planning
to go whole-hog for OS/2 development sees IBM pull back just when it was on the
verge of success, how will this same vendor feel about supporting an IBM e-commerce
initiative? Will they want to support IBM's PowerPC development, when IBM resumes
that direction again?
There is something to be said for a slow, steady drizzle instead of a series of
cloudbursts. Hopefully, IBM has finally caught on to the importance of sticking
to a single, all-encompassing message. The Java and related open-standards momentum
at IBM will hopefully be the sign of a long-term trend, instead of just the latest
cycle of thrashing.
Most recent revision: January 9, 1998
Copyright © 1998, Tom Nadeau
All Rights Reserved.