Re: The software factory (was "Don't believe the hype?") (long)

Subject: Re: The software factory (was "Don't believe the hype?") (long)
From: Andrew Plato <gilliankitty -at- yahoo -dot- com>
To: "TECHWR-L" <techwr-l -at- lists -dot- raycomm -dot- com>
Date: Wed, 10 Mar 2004 11:02:59 -0800 (PST)



"Mike O." <obie1121 -at- yahoo -dot- com> wrote in message news:231599 -at- techwr-l -dot- -dot- -dot-

> Repeating successful IT projects is like weather forecasting. If you have
> enough data points, and enough processing power, you can predict every gust
> of wind for the next year. But nobody puts that amount of effort into
> planning and predicting IT projects; therefore tech managers are often
> surprised when it rains.

This is one of the more enlightening comments I've heard on TECHWR-L in a
while.

Complex systems (like designing software and writing docs) can be highly
structured. But there are many variables. And few companies have the resources
or might to gather all the necessary data and analyze it effectively. (Which is
basically a repeat of what Mike said.)

Now, to expand on that:

Therefore, the best methods for handling complex problems is a scientific
method.

- Observe
- Hypothesize
- Test
- Draw Conclusions
- Document
- Peer Review

This method works because it does not assume the input is identical every time.
It leaves a lot of room for the individuals within the process to govern the
output. Manufacturing processes focus on standardizing inputs and outputs, so
they can work within a standardized process. Hence, car manufacturers don't
just carve cars out of raw steel, they standardize the steel into specific
formats. Then they run it through a standard process.

This works because the inputs have been standardized and hightly
compartmentalized. Steel manufacturers don't make cars. They just make raw
steel in standard formats. Body processing plants don't make engines, they just
take standardized steel and bend it into bodies. Engine plants don't make
radios, they take ... you get the idea. High standardization coupled with
tremendous compartmentalization.

So what happens when you cannot control the inputs? They are not standardized?
Or they vary widely over a period of time?

Manufacturing processes break down at that point and become more of a hindrance
than a benefit. As such, you must switch to a scientific process. Where you
accept that inputs vary and hinge the success of the process on the successful
analysis of information.

Most "intelligence gathering" systems are scientific oriented. Attempts to
turn them into "intelligence mills" has had disastrous results. (See 9/11).
Intelligence failures are typically due to data that failed to be analyzed
properly. This can be a systemic failure (the process does not look at the
correct factors) or a input failure (the data was overly standardized and in
the process lost its nuance and key details.)

Standardized processes demand standard input and produce standardized output.
If you can align all three (input, process, and output) and produce a useful
product, then manufacturing processes work and work very well.

But if any one of those points cannot be standardized, then the entire process
will not work. And in tech writing, its virtually impossible to standardize the
input and the process. Especially when you have many writers who do not
understand the input they are handling.

Andrew Plato

__________________________________
Do you Yahoo!?
Yahoo! Search - Find what you?re looking for faster
http://search.yahoo.com

^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

ROBOHELP IS THE INDUSTRY STANDARD IN HELP AUTHORING
New RoboHelp X5 includes all new features such as,
content management, multi-author support, distributed
workforce support, XML and PDF support, and much more!

Purchase new Macromedia RoboHelp X5 by March 31st and receive a
$100 mail-in rebate. http://www.ehelp.com/techwr-l

---
You are currently subscribed to techwr-l as:
archiver -at- techwr-l -dot- com
To unsubscribe send a blank email to leave-techwr-l-obscured -at- lists -dot- raycomm -dot- com
Send administrative questions to ejray -at- raycomm -dot- com -dot- Visit
http://www.raycomm.com/techwhirl/ for more resources and info.



Previous by Author: Re: not really OT: eweek article 'Popup Ads Bad for Business'
Next by Author: Law of Morons (was flashpoint of the week)
Previous by Thread: RE: The software factory (was "Don't believe the hype?") (long)
Next by Thread: Re: The software factory (was "Don't believe the hype?") (long)


What this post helpful? Share it with friends and colleagues:


Sponsored Ads