Usability and Software Fitness

Gary Lawrence Murphy garym@canada.com
21 Aug 2001 02:24:09 -0400


>>>>> "t" == tomwhore  <Tom> writes:

    t> Its not that one is slave to the other or less than the other,
    t> its that you must must must have a blend of functions such that
    t> the USER is served best, whoevere or whatever the USER of that
    t> thing is.

I still get blank stares when my design docs include use-cases for
operations, maintenance and management.  I recently watched both a
non-profit and a dot-com pour thousands into promoting their work with
give-away kits, never once thinking that they'd missed the golden
opportunity to at least collect a whack of business cards, and we've
all worked on projects where the bad designs rushed through to give
one project manager on-time/under-budget status only served to swamp
the budget of the customer-support or the QA team.

I don't use the word "users" in my design work. I prefer
"stakeholders" or "participants".  Drugs have users.  Software has
participants, only some of which are human, only some of whom enter
via the GUI. I also get blank stares for including "uninstall" as a
necessary life-cycle phase, but that's another thread (why is it every
dot-com thinks theirs will be the only product in the Universe, and
that it will live forever? ;)

Software is a system, but no system exists in isolation. It must
always be intertwined with some other system.  Not nice and neat like
nested carved ivory balls, but all blurry and twisted into Kline
bottles of interaction.  Business software, surprise surprise, sits
within the requirements spec of the system called business, which
itself is entangled in other specs for "law", "society" and so on and
so forth.  Ergo, if you expect an intellectual investment in software
to sustain a pack of salaries and shareholder profits, then, yes, its
positive effects should be expressable in ledger sheets ... and in
such a way that the IRS is not involved :)

The I Ching says, "In order to become whole, things must extend beyond
themselves, into the infinite" but I expect an equivalent translation
might say "things _will_ extend".  Even a lowly CGI on a cheapo (or
free) webhost plan co-exists with other CGIs and is accountable to the
ISP's sysadmin and the webmaster of the site as well as doing it's
little narrow task.  There are use-cases in all directions.

Both management and engineering can be too eager to define a membrane
with too small a circumference.  Maybe it's because it supports the
illusion that we can actually control something, maybe to avoid
uncertainties in the requirements or to avoid responsibilities in the
outcomes, or maybe it's just laziness.

Ok, all that said, in _my_ experience, it has not been free software
or even free services that has lit the dot-bombs.  I went through the
same rampant dot-crash landscape with the old client-server companies
in the late 80's -- it's not that they didn't sell software or that
they gave away too many services, they floundered because they failed
to apply realistic whole-system views to their _business_ systems.
They mistakenly believed the only worthwhile "users" of their business
system were the shareholders, and as they abandoned all other
stakeholders, the whole thing just starts to unravel until it falls
apart.  When the host dies, the parasite dies too. 

I've yet to put it to an experimental test ;) but I believe the trick
is not to make investors happy or make the customers happy or make the
developers happy, the trick is to make everybody _happier_.  As
someone pointed out, a good grasp of the entire business system can
succeed even when the engineering is slack because all it needs to do
is to make _more_ people happy. If it makes people happy, they will
sustain it, so the equation is only to reach that level of stakeholder
happiness that keeps the host alive. (I disagree about the total
prowess of management over engineering, since excellent marketing can
make a bad product fail very quickly)

(geez, what a way to come bounding on to a new list ;)

-- 
Gary Lawrence Murphy <garym@teledyn.com> TeleDynamics Communications Inc
Business Innovations Through Open Source Systems: http://www.teledyn.com
"Computers are useless.  They can only give you answers."(Pablo Picasso)