[comp.software-eng] controlled change

rcd@ico.isc.com (Dick Dunn) (09/29/90)

johnb@srchtec.UUCP (John Baldwin) writes:
> rh@smds.UUCP (Richard Harter) writes:

> >Many developers are uncomfortable with the notion of controlled
> >change; however...

> Its a sad but funny fact that, in fact, many developers are more
> comfortable with the notion of "uncontrolled change."  All too many
> projects (even well-managed ones!) are planned and completed with the
> view that "once its done, its done."

I'd dispute that a project is well-managed if it takes that view.

> How many of you can remember a real project that was completed and
> never changed?  [I'll admit, there must be a few.]

I have seen projects with that characteristic.  They all have one other
characteristic in common: They were never used!  For one reason or another,
they were shelved between when they were "finished" (i.e., first real
release done) and when they were scheduled to go into use.

In fact, we have a sort of saying around here related to that effect...
	"How did people like the xyzzy project."
	"No user-reported problems or change requests."
	"Really?!?  Why did it get cancelled?"

Software *always* changes.  If it doesn't work, it must be changed so it
does work.  If it does work, it will be changed to do something different.
-- 
Dick Dunn     rcd@ico.isc.com -or- ico!rcd       Boulder, CO   (303)449-2870
   ...Worst-case analysis must never begin with "No one would ever want..."