[CLUE-Talk] Quality comments

[Kevin Cullis]

Lynn Danielson wrote:
> 
> Grant Johnson wrote:
> > > > counter-productive.  I was in a CMM level 3 organization.  In order to
> > > > achieve this level, a full 1/4 of the staff did nothing but administer the
> > > > program, and the other 3/4 were micro-managed into hating their lives and
> > >
> > >Yes, even I hate it when someone finds my mistakes, and depending on how
> > >they approach me, I'll either hit them with the nerf arrows for pointing
> >
> > It isn't the pointing out of mistakes.  It is the paper trail and jumping
> > through hoops that takes so much more time that doing the job that you no
> > longer have time to do the job right.
> 
> It's a given that most of us are interested in computing -- not
> documentation and accounting.  While I can intellectually see the value
> of a quality system, I have yet to see one that gave me a warm fuzzy.
> "Ooooh, I just love this quality system it makes my like my job so much
> better" is not the type of comment I've ever heard from anyone who's been
> subjected to one.  The extent to which a quality program can be minimally
> intrusive and low maintenance will be directly related to its acceptance
> and success.  From my experience and others, most quality control systems
> fail miserably at this.

And there's the rub: no one wants to do better or think that they are
the best when really they're not.  Take Linux for example (you knew that
I'd get around to it sometime ;-) ), if it weren't for Linux and it's
fast growth, where would M$ be today without the Linux influence? Still
raking in the bucks with NO competition and we'd be paying even more
bucks for their lousy products. It is only when someone says "I hate
this level (definition) of quality, I'm gonna do something about it"
that something is done, and that's EXACTLY what Linus did with Linux. 
HE changed the paradigm of software development (as well as some of the
others in the Linux community) and has caused competition to occur
within the OS community.

Most people don't like change, until there is a paradigm shift and YOUR
job gets affected, THEN you'll change.  With Linux gaining ground
everyday, how do you think M$ is reacting?  They're smart, so they'll
change and they have.  Now whether the change reaps the benefits they
want is left to be seen, and there, too, is the rub: we won't know until
down the road to see if they succeed or fail.

Since most of the comments about quality of late have been questionable
about process improvement and its benefits, let me ask all of the Linux
gurus out there some questions: what if you didn't have the
/var/log/messages DOCUMENTATION file, what would you do to find problems
out on your systems?  Who designed the /var/log/messages stuff and why? 
Prior to /var/log/messages, what did the users have to find out what
happened to their computer, i.e. BEFORE someone AUTOMATED this
function?  You can't automate something until the process is stable.

So, the idea for process improvement is there, the question remains:
what are we/I/you going to do with this information?  Some of us won't
do a thing "until we/I/you have to."  Others, "eh, I'll try it" and
still others will say "Ooooooh, I can see how my work DOES affect others
within my organization."  True, it's a balancing act of performance
between the IT department and it's internal customers with end result
that your EXTERNAL customers, those who pay your bills, keep coming
back.  Too often, one extreme, the internal customers (the inmates) run
the IT department) or the other, the IT department runs the company,
comes down to becoming a power stuggle (what I want) rather than a
performance stuggle (what does my EXTERNAL customer want).

> subjected to one.  The extent to which a quality program can be minimally
> intrusive and low maintenance will be directly related to its acceptance
> and success.  From my experience and others, most quality control systems
> fail miserably at this.

And this is what takes the balancing act. The key point is that most of
us, including me sometimes, don't like to be bothered with this stuff,
but once I've seen some of it's benefits beyond "it'll NEVER work"
rather than HOW MUCH will it work, I was convinced there was good reason
to do process improvement and to learn the skills needed to improve.

Years ago after Dr W. Edwards Deming had his phone ringing off the hook
by MAJOR Fortune 500 companies to help them (mostly Detroit) with
beating back the Japanese (or trying to catch up to their level of
quality).  He asked one question: would the CEO of that company be in
the same meeting room as his underlings?  Yes he would.  Deming showed
up and was introduced by the CEO to his hundreds of underlings in the
meeting room that Dr. Deming was to teach everyone about quality.  The
CEO left the room with Deming picking up his stuff and walking right
behind him heading for the door.  The CEO said "Where are you going?" to
which Deming replied "If you don't know what I'm teaching them about
improving quality, how will you be able to COMMUNICATE and provide
DIRECTION with them CORRECTLY?!"

It took them almost a year to get him back to help them.

The authors of "Enlighten Leadership" wrote the book for the C?Os of the
world, but you know who bought them?  The middle managers! Why? Because
the higher ups rest on their laurels (other than GEs CEO Welch) and have
"arrived."  Too often, this is the case and I fight it all the time.

I didn't mean to go on an ESR type rant/rave, but I had to.  ;-D

Til next time.

Kevin
Linux Evangelist for Short Sighted Suits IS Making Everyone Rethink
Everything ;-)