[lug] [CLUE-Talk] Mundie and the GPL (fwd)

J. Wayde Allen wallen at its.bldrdoc.gov
Mon May 21 09:05:52 MDT 2001

An interesting point of view on the Open Source versus Proprietary
software debate.

- Wayde
  (wallen at its.bldrdoc.gov)

---------- Forwarded message ----------
Date: Sat, 19 May 2001 10:31:14 -0700
From: Grant Johnson <grant at amadensor.com>
Reply-To: clue-talk at clue.denver.co.us
To: letters at news.com
Subject: [CLUE-Talk] Mundie and the GPL

What Mr. Mundie does not understand is where the money REALLY is.  Yes,
small numbers of people can large sums of money with closed source,
however, companies spend far more money on the labor to implement and
maintain that software than they do on the software itself.  This is
where the real money is.  These many individuals making $70,000 per year
add up to quite a lot more money than those few at the large software
vendors making their millions.

Yes, software drives the economy, but the vendors in reality are a small
portion of it.  Then the question becomes more "How do we get more for
our labor" than "How do we keep our purchase costs down."

The GPL, while not sensible for a company producing shrink-wrapped
boxes, does make sense for companies developing in-house systems, as
well and services and consulting firms.  The community effort behind the
software spreads the effort to reduce costs to individual companies,
while the openness ensures interpretability, speed of development,
maintainability, and the future of support for the software.  Open
source is more supportable.  Imagine, when Word Perfect was king.  It
held a near monopoly position for word processing software.  If I had
told you then that it would be bought out by Corel who would then go out
of business, you would have never believed it.  Now you see that this is
a real possibility.  What happens to our support now, if we have, over
the years, standardized on Word Perfect, and they do go out of
business?  If this were an open source product, someone else could pick
it up if it were profitable enough, and if not, we could support it
ourselves if it was important enough.

The following is an article I wrote for a local open source user group
news letter.  It explains that why, now that the cost of distribution is
low, the GPL makes sense.  What Mr. Mundie fails to realize is that
intellectual property, unlike other property, is not exclusive in its
use.  Both he and I can use the same piece of information at the same
time, unlike a car, where we cannot both drive at the same time,
especially since he is on the West Coast and I am in Colorado.  We, as a
society, need to reassess how we trade in this, our most valuable
product, information.

There are only two good reasons to write software:  To wrap it up in a
pretty box covered in plastic and sell it, or to use it.  Sure, one
might write software to print it out as birdcage liner, or out of sheer
boredom, but I said GOOD reasons.

Let us examine wrapping it up in pretty boxes.  In order for this to be
successful, the software must be attractive to the widest possible
audience.  This involves having every feature anyone could possibly
want, as well as a great deal of marketing.  

This has downfalls.  First, the people making it never actually have to
use it, so they have no clue what they may be subjecting their buying
public to.  Second, this is driven by new releases for profit, whether
they are really needed or not.  Third, this upgrade driving profit
driving development means that the new version must be made mandatory,
if not because it has great new features, then because it obsoletes, or
breaks the old version (This has been most noticeable in office suites.)

And now for using it.  In order for this to be successful, the new
software must save enough expense to justify the expense of making it.  

This also has its problems.  Not everyone who needs software is a
programmer or knows one willing to help.  Often interfaces will make
sense only to those who made it, and not appeal to the general public. 
Some projects, in isolation, are not cost effective.  If everyone who
wanted a word processor had to hire a programmer to make one, we would
all still be using typewriters.

If all of the individuals who need a word processor pooled their
resources, it would be easy to find the funds to hire programmers to
make one.  This community effort is very similar to communism (the
economic system, not the often associated governments.)  This has been
proven over time to work very poorly, due to class structures that
develop, and general human greed.  

What is the answer then?  

Open source to the rescue.  The project, if small enough, is cost
effective.  A simple text editor, for a company with a lot of typists,
is a cost effective investment.  They then release this code, GPL, to
the public.  Since this code was developed for the purpose of using it,
rather than for selling it, the cost of releasing it is negligible, and
the company still gets the benefit of using their text editor.  Now
comes company B.  They need a more advanced system, with bold and
italics, etc.  They cannot justify the expense of making this, however,
the expense of adding these features to the existing text editor aren't
so bad.  They add these features, and share them with everyone,
including those that originally made the text editor.

So, if you, as a programmer, write software for your company to use
rather than sell (the latest figures I have seen place this at 80%-90%
of all programmers) then you can only benefit from releasing what you
develop under the GPL, leeching improvements others make to it, and
taking software that is close to what you need and improving it, adding
Which of these is likely to produce better, more efficient software? 
One tries to cover every function anyone could ever want, but never
actually uses the software themselves.  The other makes just what they
want, but they do so to use this software.  The open source may be
missing features you want, but the features which are present are more
likely to work.  Which do you want, a car with no air conditioning, but
reliable, or a car with air conditioning, alarm, neon lights, three
sunroofs, and a built in coffee maker, but with an engine that catches
on fire every time you open the trunk?

Some claim no one can make a living from GPL software.  My making a
living is not dependent on my source not being borrowed.  I am brought
in to fix a specific problem.  Sometimes this involves writing new
software.  Sometimes it is implementing an off the shelf solution.  Most
of the time it is in between.  Everything I have ever done has involved
some code from somewhere else, at least in the OS.  Whether the client
buys this, or whether I download it from someone else, I get paid the
same.  The only programmers who could be hurt by open source are those
writing shrink-wrapped software.  This is less than 20%.  The extra
money made available by sharing effort would more than balance this
out.  Those programmers would just start doing what the rest of us do. 
Of course the CEO types of shrink-wrap corporations do stand to lose a
lot, but no one else.

In the software world, there have been two warring camps.  One saying
that all proprietary software is evil, and just a corporate effort to
take over our lives, making us all slaves in a giant machine, eventually
to be ground into Soylent Green.  The other saying the open source is
Marxist, Communist, and generally un-American, destroying the goodness
of apple pie, and the possibility to make a living as a programmer at
the same time.

Newsflash:  You are BOTH wrong.  Like so many things in life, the
circumstances dictate the correct solution.  A hammer is not always the
right tool, but neither is a saw.
CLUE-Talk mailing list
CLUE-Talk at clue.denver.co.us

More information about the LUG mailing list