Re: bias

On Sat, 8 Jun 2002, Francois Gouget wrote:

> On Fri, 7 Jun 2002, Deven T. Corzine wrote:
> [...]
> > > First, most skilled computer professionals do not make a living off
> > > commercial software: most make their money off in-house support and
> > > in-house development.
> >
> > Be that as it may, end-users still have a need for software products, and
> > the GPL definitely threatens the development of end-user products that
> > aren't worthwhile to a company to pay for, even if millions of home users
> > would happily pay...
> 
> I don't understand your reasoning.
>  * if the development of such a product is not worthwhile in the first
> place, then companies are not going to develop it anyway

Worthwhile to whom?  The product may be worthwhile to millions of end users 
who are only able to pay, say, $100 for it.  It may, however, cost millions 
of dollars to create the software.  It won't be worth millions of dollars 
to any particular organization to ensure that it will be created, and if 
there's no profit incentive, nboody will risk the money to make it.  And 
without the ability to coerce payment from users (which the GPL forbids), 
there's no way to recoup the costs of development, much less make a profit.

>  * if it is worthwhile, then companies are going to seize the opportunity

What opportunity?  If the product is developed under the GPL, they'll sell 
a few copies, and most people will get it free from a buddy.  Meanwhile, 
the company that fronted millions of dollars for the development never sees 
a fraction of that money coming back.

This is an opportunity to throw money away on development.  Real enticing.

>  * if a *GPL product does really make it not worthwhile for a company to
> develop a given product (which may happen but certainly is not the
> general case), then the millions of home users are not left out in the
> cold since they now have an open-source product to use

My point is that there's little incentive to create that *GPL product 
itself, in the first place.  That's why it tends to be done mostly by 
volunteer effort.

Also, you can't count Mozilla and OpenOffice.  They're only free because of 
Microsoft's stranglehold on the marketplace -- huge companies fronted TONS 
of money to develop these products and deliberately "dumped" them on the 
market for free to break Microsoft's grip on the industry.  This is pure 
capitalism, and a major gamble on the part of AOL/Netscape and Sun.  They 
expect the value of a more competitive marketplace to exceed the cost of 
giving away these valuable, expensive products.

>  * if a *GPL product has such an effect, then the same product with the
> same level of functionality under an X11 or *BSD license would have the
> same effect on the market. And if proprietary extensions thrive it's
> because the X11/*BSD product is so bad that it is unusable for most of
> these millions of home users. Do you believe that open-source projects
> cannot get any better? And what happens once they do get better (which
> is inevitable)? By your reasonning they will kill all proprietary
> offshots again.

I think open projects will tend to get better.  Slowly.  Many users are 
impatient waiting for the natural pace of these things, and want results 
quicker.  Push come to shove, they'd even be willing to pay for it, but 
they just don't want to.  They'll be happy to leech off the public project 
indefinitely, but rarely choose to contribute for ideological reasons.

> So I really don't see a problem and I don't see how this is specific to
> *GPL projects.

This is specific to the GPL (not as much the LGPL) specifically because of 
the coercive, viral, proprietary nature of the GPL.

> So what is the alternative anyway?
> Make all open-source and public domain development illegal?

This is a ridiculous answer, as you know.  Non-proprietary code is fine -- 
it can always serve as a basis for newer proprietary offshoots.

The GPL codebase is a particular danger because it's a proprietary codebase 
that is ALWAYS "dumping" on the market, by definition.  Many people thought 
it was unfair competition for Microsoft to undercut Netscape by making the 
Internet Explorer browser free of charge.  It's classic "dumping".  It was 
done (by Microsoft) for the sole purpose of breaking Netscape's monopoly on 
the browser market, and it worked.  It's also considered an antitrust 
violation for a monopoly player to do this, because it squashes the market.

In response to Microsoft "dumping" Internet Explorer, Netscape decided not 
only to "dump" their browser on the market free of charge, but ultimately 
to make it a free software project as well.  Sun later decided to free the 
StarOffice code (which they paid millions to buy), and even under the GPL, 
as another "dumping" maneuver to try to break Microsoft's Office monopoly.

As a user, I hope Mozilla and OpenOffice succeed in breaking Microsoft's 
strangehold on the marketplace.  They seem to have a chance of doing so.  
But I don't delude myself into thinking they're shining examples of the 
inherent superiority of open source software.  Those projects would NEVER 
have gotten all this development and attention without SERIOUS funding from 
enormous companies willing to take an enormous risk to counter an enormous 
threat.  The big players are jockeying for position, and we're watching 
from the sidelines.

Netscape had something like 300 developers working on Mozilla full-time 
PLUS volunteer contributions, and it STILL took over 4 years to get a 1.0 
release.  That's about 2.5 million paid man-hours of development work -- 
just how long do you suppose it would take for a project to amass millions 
of hours of volunteer labor?

The GPL envisions some sort of utopia where programmers work for the love 
of the craft and the whole world benefits from all the code anyone writes.  
This sounds wonderful in theory.  In practice, like most utopian schemes, 
it seems likely to be fatally flawed.  Can you prove that it's not?

Deven