Re: Not an awk question

[Michael JasonSmith <MPJ17@HIDDEN>]

On Fri, 2002-11-08 at 23:35, Zane Gilmore wrote:
> There is an entire development methodology (whose name escapes me at the
> moment) that makes use of that very phenomenon.
Apple found this out to their benefit when they were making the
Macintosh.  So here is a parable from Apple.

---==---

Once upon a time, before the Macintosh was released, the documentation
team was working alongside the coders, writing the API documentation as
the features were added to the GUI toolkit.  The coders resented these
lightweights coming in and asking them questions about the API, as it
slowed down the coding.

One day a member of the doc' team walked in and asked the coder to
explain the API for the list-widget.  The coder tried, but was having a
lot of difficulty.  The coder tried a number of different ways of
explaining the API, but none of them were very satisfactory.  In the end
the coder asked the documenter to come back the next morning: "I'll have
figured out how to explain it by then."

Over-night the coder rewrote the *entire* list-widget API, the
documenter came in and received a clear, precise explanation about how
the list-widget worked.  The documentation was clear, and the API was
easy to use.  The coder realised why they had the documentation team
asking them questions, and didn't resent the interruptions any more.

---==---

I read the full version of this story on the Apple Macintosh history
site at Stanford, but I have forgotten the URL.
-- 
Michael JasonSmith      http://www.cosc.canterbury.ac.nz/~mpj17/