[OT] What is Open Source? (was Re: ANN: Twisted 0.16.0...)

[Stephen J. Turnbull]

>>>>> "Laura" == Laura Creighton <lac at strakt.com> writes:

> >>>>> "Laura" == Laura Creighton <lac at strakt.com> writes:
> 
>     Laura> People here seem to be thinking that the GPL is there to
>     Laura> protect creators rights to their software.
> 
> I'm pretty sure that that is what rms has in mind, as long as both
> "creators" and "software" are inclusive collective nouns.
> 
> The consumer protection argument is the OSI line (a special case
> thereof), and rms don't have much truck with anything so economical
> and relativist.

    Laura> 1. Why did you remove my prior line, 'as a thought
    Laura> experiment?'

I apologize for the rudeness, none was intended.  I did not, and still
don't, see that its presence or absence has any effect on the
semantics of the discussion.  So I took it out.

    Laura> I put it in in order to avoid having this conversation
    Laura> right now?

But it didn't.  I don't see how the GPL is about consumer protection.

Not as "original intent" (you _did_ use the word "intended"), and not
as a "thought experiment" in social engineering.  I planned to address
both, but I see I completely missed the mark on the "thought
experiment" (and further confused the issue by mentioning rms again).
Again, I'm sorry for creating misunderstanding.

There are plenty of ways to address the goal of consumer protection,
many of which might very well be more profitable (eg, the US DOD's
second source policy) for the developer of the software under license
than the GPL.  None of these are satisfactory substitutes for the GPL
(or free software in general) because they violate freedom.  Another
example: the Aladdin Free Public License has all the benefits of the
GPL in protecting customers from broken software: you can always hire
someone to fix your broken Ghostscript.  You just can't redistribute
the fixed version for money, and thus amortize the cost.

But you don't want to go there, do you?  rms definitely doesn't.

If one is _only_ interested in consumer protection, there are plenty
of very attractive alternatives to the GPL that could be imagined.  As
you point out, some might have to be enforced by changes in IP law in
practice, but most could (in theory) be achieved simply by customers
insisting on a reasonable contract.  (The OSI line again.  "You don't
need to abolish IP or to demand the GPL; there are other ways to
protect yourself.  But you should protect yourself by insisting on
access to source.  Preferably by using free software, which benefits
all software users, usually without hurting you.")

    Laura> 2. I learned this from RMS himself, when he was living in
    Laura> my house about 15 years ago.  Look up 'hoarder' in his
    Laura> writings.

I've read them.  I understand the concept.

My mother doesn't.  She will never notice how much software has been
"hoarded" by Microsoft, only how much has been delivered at remarkably
low prices.  My mother has no need or desire for a copy of the source
herself.  She recognizes the usefulness of _someone_ having access to
source, but "second-sourcing" or "source escrow" would be plenty to
protect that wily consumer in her own opinion.

My point is that rms developed the GPL to protect/enforce sharing of
software ideas, not to protect consumers' economic interests in
working software.  The protection of consumers is a very desirable
side effect, but according to rms _economic benefits in general must
never be considered on a par with software freedom_.  The GPL wasn't
intended to protect consumers, and it is not tuned to doing that
effectively, since it surely results in less software being developed.
There's a very good reason for rms's denial of consumer sovereignty:
it is easy to show under the standard microeconomic models that "a
little bit of IP in the right places" will surely benefit consumers
relative to "no IP," and the GPL imposes _more_ frictions than "no IP"
does.

So I don't think an argument starting from consumer protection will
lead to the GPL, or anything that looks like it, in many important
cases.  rms certainly doesn't.  He says so every time he's asked to
comment on "open source," and he's personally asked me not to
formalize, not to measure, not to mention, and most especially not to
publish what "is easy to show...".[1]

Thus my point that the GPL (and software freedom in general) is
primarily about protecting the freedom of developers to develop.

    Laura> I thought compelling people was a worse sin than software
    Laura> hoarding.  [...]  RMS does not see things this way at all.
    Laura> He thinks that software hoarding is a worse sin than
    Laura> compelling,

I've seen him write similar things, also the moral equation of
intellectual property to human slavery.  But I don't understand what
he means by this, as I would deduce a need to close the "ASP loophole"
in the GPL.  Eg with a "must publish `deployed' modifications" clause
a la the Apple Public Source License.  But he firmly rejects that.[2]

By the way, there's a small technical error in your consumer
protection argument.  You wrote

    Laura> The thing is you will be able to tell (or hire somebody to
    Laura> tell you) before you buy

There's nothing in the GPL that requires that you get source _before_
you buy the software.  Cygnus and The Kompany both have made money on
GPLed software for which the source is not public.  It just happens
that the customers had no interest in redistributing source
themselves.  In fact, it seems plausible that the GPL was helpful in
convincing some customers not to redistribute the source, as it very
likely would have quickly ended up in the hands of a competitor.

I doubt this is all that important in practice yet---most free
software seems to be public source as well as open source---but I
suspect it will grow over time, especially in market segments
developing for ASP-like applications.[3]


Footnotes: 
[1]  The "standard microeconomic models" are far more applicable than
most free software advocates would like to believe.  Showing what the
limits and appropriate modifications are is a project I guess I should
dust off and get published.

[2]  For reasons I understand, I think, and agree with.  So evidently
I don't understand the hoarding v. compelling and IP v. slavery ideas.

[3]  Though it may not.  The Kompany now avoids the GPL where
possible, as they found that bad PR from not publishing source to
GPLed software is actually more unpleasant than bad PR from
distributing software under proprietary licenses.

-- 
Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN