[CLUE-Talk] CMM and Process Improvements

[Kevin Cullis]

Jeff,

I know I'm not your typical "developer" but I'll answer your questions
anyway just for comparison sake.

Jeffery Cann wrote:
> For example:
> 
> 1.  What kind of software are you producing? (web, database, client, etc.)

A process is a process, however it is defined, be it simple or
difficult, large or small, short or long. The ultimate question is: are
you getting better (cheaper, faster, better quality, and yes, you can
get all three, over time) at producing your product or service to your
customer?

> 2.  Is your software written for government agencies?

See above

> 3.  How many folks are in the process.

The more people, the more difficulty, but if you've got a small group of
PHBs, it the same.

> 4.  What level you are working at within CMM.

See below at 5.

> 5.  How was the transition to CMM?

This is the most difficult.  Why?  Because it requires people to change
and most people don't want to BE changed!  As most people view this
issue (same with process improvement in general), it should NOT be
looked at as a pass or fail, but one of how much better have you gotten
over time?  Too often I hear people state that an endeavor "fails" and
then laugh at the effort to improve when it doesn't reach expectations.
The true question is: how much have you improved?  Is the rate of
improvement fast enough for your customer? Competition? Your company?

> 6.  Why is it one of the best process models?

If you look at the level 5 versus the level 2, what's the difference? 
Clearly defined processes with concise metrics and everyone knows where
they fit within the picture. It's no different than a manufacturing
plant, the principles don't change!

> 
> others?

Jeff, the single most disturbing thing about this issue of improvement
is: people are scared to try and want only success without little risk. 
If 10 companies in Colorado are doing CMM out of 100, the only way for
the next 10 companies to try CMM have only a few choices to follow: hire
(away the learned which will increase the cost over non CMM talent) or
pay to learn (at a much less cost than hiring) what the other 10 are
doing, DIY (Do It Yourself) with minimal costs but will/may take longer,
or do nothing at all.

Just like I was introduced to Linux and have become a convert, so those
of us who have worked with process improvement don't want to go back to
an older, less effective/efficient way of doing business.  Once you're
hooked, it's hard to go back, it now becomes a matter of getting others
hooked into seeing the light.

Kevin