Only after you make your system usable, reliable, and
efficient, should you consider making it fast. In fact, when you make
your system usable, reliable, and efficient, it’s going to be faster
by default, but not due to any actual increase in system speed. The
increase in speed will come because you spend less time trying to
accomplish a task and fixing system errors. You might be surprised
if you monitor the time lost due to various inefficiencies today.
However, the fact remains that you can usually speed your system up
once you have all of the other problems under control.
The problem for many people is deciding what to first.
There are many sources of application time wasters. The “Tuning Your
Applications” section of Chapter 1 provides a good starting point.
Optimizing Windows first and then moving on to individual applications
is a good plan. By removing the waste from Windows, you gain an overall
speed increase for all applications. In some cases, the speed increase
might not be enough, so you also need to look at individual applications.
Optimizing applications isn’t the ending point for the
optimization process. When a user opens too many applications, the
system will still slow down and have reliability problems. Part of
the optimization process comes in the form of user training. It’s
usually not a good idea to keep five or six main applications open
at once unless they’re actually in use. Many users will open every
application they plan to use for the day and keep them open all day
long. The result is an overburdened system. Closing and opening applications
as needed results in a system that’s less confusing to use, performs
faster, and makes better use of resources.
Consider the scenario of a simple download. The application
in question can open several threads to perform multiple download
tasks at once. This process uses network bandwidth more efficiently
and to a fuller extent, which means the application completes the
download more quickly. However, because the user has five or six other
applications open — applications that aren’t even in use — the download
application doesn’t have access to the required memory, so the download
languishes. The user naturally blames the slow download application,
rather than the true source of the problem: lack of memory to complete
the task.