Subject: Re: Free Software vs. Open Source
From: Ben_Tilly@trepp.com
Date: Tue, 12 Feb 2002 19:56:50 -0500


Larry Augustine wrote:
> I have changed the subject of this thread to something that is probably
> more descriptive.
>
> I have not been able to discern any difference between Open Source and
> Free Software with regards to the end result: source code being "free"
> in the sense that the FSF defines "free".  Both communities seem to have
> the same goal.

Indeed the open source definition evolved directly out of the Debian
community's attempt to define what free software is.

But the communities in question actually don't share that goal.  While
the OSI may have been set up to generate the production of more open
source software, many people who rest under the open source umbrella are
not wedded to the idea that open source should be everywhere.  Indeed
part of the *point* of the OSI was to craft a message that results in
what free software folks want (good free software) from people who are
not about to agree with RMS.

For example Larry Wall has gone to lengths with Perl to make it clear
that he thinks it is fine for people to take what he has given away and
use it within proprietary products they resell.  Indeed Larry's beliefs
on this are strong enough that the Perl community has had to go well out
of its way to duplicate work that exists in the GPLed world just to
avoid being bound by that license.

You may think that this point is an unimportant nitpick.  But people
don't do a lot of work because of a point unless they care about the
point.  So no matter how minor the point looks to you, actions show that
developers care.  (Ask Simon Cozens how much extra work is involved...)

> However, the Open Source community seems to believe two things which RMS
> and the Free Software community do not seem to believe:
>
>    1. The word "free" is detrimental to achieving the goal because
> most people seeing it applied in the context of software think "zero
> cost", not "freedom."  By using a different term, it may be easier to
> achieve the end goal.
>
>    2. There may be reasons for supporting Free Software other than
> a moral belief that all software should be free.  The Open Source
> community is willing to be flexible in the rationale behind Free
> Software and is accepting of pro Free Software arguments based on
> economic or other rationale.  RMS and the Free Software community do not
> seem to allow for these other arguments.  Again, by allowing for a
> different rationale, it may be easier to achieve the end goal.

I don't think that the first is a point of major disagreement, I have
certainly heard free software folks complaining about the ambiguity of
the word "free".  As for the second, well RMS thinks that there is
nothing wrong with having people who disagree with him write free
software because they need to use something that is GPLed badly enough
to apply the GPL to their own work.  While he prefers that people agree
with his agenda, it isn't necessary that they do so.

The point of major disagreement is that the open source message is more
accepting of people whose opinion on free software is much more shades
of grey, shading all of the way to black.  There are theories for the
value of free software for developers, customers, and businesses.  It is
possible to accept any combination of these messages.  It is possible
(indeed ESR has encouraged this) to believe that open source is good in
some situations and questionable in others.

In other words the OSI set out to craft a message that would get people
to create free software without their having to accept the free software
dogma - and succeeded.  As a result there are lots of beliefs tolerated
in the open source community that are anathema to the goals of the FSF.

> I believe that RMS and the FSF have slowed the acceptance of Free
> Software by their unwillingness to compromise on these two points.  I
> believe that adopting the terminology and arguments presented by the OSI
> would accelerate the acceptance of Free Software.

All of the detailed listening to and thought about the open source
message will not convince you of the free software message.  Its success
at getting people to contribute to and write more free software is good.
It has not had the same success ratio on convincing people of the FSFs
goals.  And I don't think it will either.

IBM has been convinced that it is in their interests to give lots of
free software away.  IBM has not been convinced that there is anything
wrong with selling software.

Apple has been convinced to build a new operating system on top of an
open source one, and to give away things related to the open one.  Apple
has not been convinced to give away anything that they think gives them
a competitive advantage.

Sony has been convinced that it can use these giveaways to compete in
the PC market without having to write its own OS and applications.
Sony's attitudes about control of intellectual property..well let's
just say that they are a big bad Japanese corporation which isn't bound
by Western attitudes about what is fair.

Hollywood has been convinced to contribute all sorts of nice graphics
code to make it easier for them to use Linux rendering farms.
Hollywood's opinion on intellectual property..well let's just say that
they don't even have the wimpy excuse that Sony does of coming from a
different cultural background.

These are all successes of the open source movement.  But if all we
ever have are successes like these, then we will wind up at some fairly
stable balance between pressures for free and proprietary software.  In
which case we _never_ achieve the goals of the FSF.  We may improve on
the past, but only to a point.  As evidence I offer the fact that
several years on this list has left me firmly convinced that there are
many situations where I do not believe it is economically wise to open
source software.

To achieve RMS' dream we need qualitatively different arguments.  RMS
recognizes this.  I just think he draws the wrong conclusion from it.
His apparent conclusion is that it is important to distance himself from
the open source movement by rejecting what it says.  Instead I agree
with Russell etc that the free software message can stand on top of the
open source one.  Different but additional, not conflicting.

Here is a possible way for free software to position itself that shows
what I mean:

  The open source movement limits itself to arguments for free
  software that do not cause any controversy.  We in the free
  software movement believe in free software for stronger
  reasons, and are compelled to question beliefs that others
  take for granted.

  The open source movement identifies many basic tasks that people
  want to do with their software.  They find great value in
  allowing people to do them freely, and great value in people
  insisting on having the right to do these things.  This basic
  list of tasks includes running, copying, distributing, studying,
  changing and improving the software.  We agree with them on this.

  Where we go beyond is that we do not accept that anyone has any
  right to DENY these privileges to others.  Therefore anyone who
  tries is wrongfully taking something of value from others.  This
  is a form of theft.  You shouldn't steal in this way, not only
  because you might be rewarded for not stealing, but because theft
  is wrong...

I believe that this line of reasoning correctly represents the OSI, the
FSF foundation, and allows free software proponents to benefit from the
work of open source advocates.  It will not please people who accept the
open source arguments but not the free software philosophy, but at least
it clarifies where the disagreement is.

Cheers,
Ben