Taxonomy of software anomalies (was: Lock, Freeze, Hang, Crash?)

[Paul Sholar <pks -at- GENSYM -dot- COM>]

Let's assume that Michael Collier's original post was not in jest, and that
he's addressing an audience who uses commercial, multi-platform, application
software products.

I agree with Valarie Tassari, and I also think that Robert Plamondon's and
Beverly Park's attempts to provide a taxonomy of buggy software behavior
are doomed.  These patterns of observed anomalies depend on the kind application
 you run, which particular product you run, which operating system or operating
environment you are running the application on, and even which programming
language the application was originally written in.

Noticing combinations of symptoms such as these IS useful to a software
engineer who is gathering information about an apparent software bug.
Attempting to describe anomalous behavior patterns in software documentation
or training implies that the user can expect to see these anomalies in the
product and that the user is expected to bring a software engineer's
point-of-view to his/her use of the product.

I wouldn't put any such description in a software product's users manual or
training course.  Instead, describe how users should report problems in using
the software, and ask your company's engineers to help your support staff
ask specific questions about reports of anomalous product behavior.


Regards,

Paul Sholar (  pks -at- gensym -dot- com  )
Sr. Technical Writer
Gensym Corporation
Cambridge, Massachusetts  USA

DISCLAIMER:  Not speaking for Gensym Corporation