🚀 go-pugleaf

OK, here is my metric for reusability of software.  The unit of measurement
is "time taken to reuse software successfully".  The way that we obtain the
metric is to take a group of "typical" programmers, assign them all an
identical set of tasks, and meausure the time it takes them to carry out
the tasks.  The tasks will all be of the form "use this package to build
a program to do XXX".

There are several caveats.  The time that it takes to learn a package
can be considerable.  We might prefer a system that has a high learning
curver but in which experts are very productive to a system that is
easy to learn but doesn't provide much advantage once you learn it.
Thus, we need to measure the time it takes an expert to use the system
separately from the time it takes a new-comer.  Also, we need to have lots
of different kinds of tasks, because most software can be used for a wide
variety of things, but it does some things very well and other things
less well.

This is how people measure the quality of user interfaces, by the way.
It seems to me that it will be the only accurate way to measure the
reusability of software.

Ralph Johnson -- University of Illinois at Urbana-Champaign