Failure to acquire and maintain accurate requirements is a leading -- if
not *the* leading -- cause of software project failure.
Writing software is hard even when you do have precise, stable requirements.
It gets much harder and takes much longer when there's imprecision and/or
instability in those requirements (the two tend to go hand-in-hand,
unsurprisingly.)
It gets harder still when the requirements contain fundamental -- but
subtle - contradictions, since a lot of product must be reworked or
discarded, which usually entails demoralizing regret and frustration.
Needless to say, a successful software effort is impossible in the case
where fundamental contradictions in requirements won't budge because
there's an unresolvable stand-off among the users, or between the users
and the sponsor. The very frequent failure of software projects in
business environments probably owes much to fundamental differences of
interest between labor and management, for example -- which is seen at its
starkest when employees want to keep their jobs but managers want to
automate employees. (In case you want to know why I bowed out of *that*
arena as well.)
I will be relying on the judgment of Dick and Jasper for accuracy in
requirements. When others appear up to speed on WG in its current form,
I'll certainly be interested. Others here are, in the meantime,
encouraged to point out contradictions in the requirements, but only where
they can make an air-tight case for logical inconsistencies that would
make a WG cognitive linguistic model fail.
Regards,
Michael Turner
[log in to unmask]
+81 90 5203 8682
www.tamaryokan.com
"Love does not consist in gazing at each other, but in looking outward
together in the same direction." -- Antoine de Saint-Exupery
|