Re: [tlug] Yes! Another argument about the GPL! You knew you wanted it....

[Curt Sampson <cjs@example.com>]

On 2009-08-11 18:06 +0100 (Tue), Lars Kotthoff wrote:

> I'm simply assuming the worst case, taking human selfishness into account.

That seems to me like a slightly odd assumption on which to work, given
that by releasing your software under a free license, you are offering
proof that that "worst case" is not always going to happen. Further, we
have plenty of evidence (and had long before the GPL) that the worst
case frequently does not happen.

> I'm not demanding profit for doing my work!...
> If you want to make a profit with the stuff I've done however, I
> think it's only fair for me to get a share of that.

This is a contradiction. However, let me be more precise here: you are
demanding payment for certain uses of your work. Specifically, should I
chose to modify your program and distribute it to others, I must give
you (and others) certain fruits of my labour that I would not have to
give you if I wrote code that did not interact with your software in
this way.

Note that I'm not saying that this is a bad thing; you're perfectly free
to do this, and welcome to do so, in my point of view. I just want you
to admit to what you're doing.


> > But, as Stephen points out, does this mean that the public has to pay
> > twice for your work in order to be able to do whatever they want with it?
> 
> That's exactly what that would avoid -- you fund it once, and then the
> work (and its derivatives) are freely available.

I derive and distribute a work from yours, I may pay twice: once through
taxes to pay you to create the work in the first place, and potentially
once again in lost revenue when I have to give away my additional work
for free. ("Free" as in "gratis," not freedom.)

> That's not the point. You can always look back on something that
> happened long ago and say "that was a good/bad decision". At the time
> you're releasing your software you have to decide what license to use
> and usually you don't have an idea what impact it will have.

Yes, but you can reason about the potential impacts, based on the nature
of the software and what's happened with software of that nature in the
past. Which I do. When I want something to be free (as in "libre"),
I use the MIT license or put the work into the public domain. When I
intend (or feel I may intend, at some point) to generate income from the
software, I use a license that restricts the freedom of others to do
so, such as the GPL. (I note you didn't respond to my point about the
GPL being, to the original author, more "business-friendly" in terms of
restricting potential competition than licenses that grant more broader
rights.)

> I didn't say that. My point is actually similar to yours -- just
> because corporations pay taxes doesn't mean that they've funded the
> research they're profiting from.

If the research was funded by taxes, I fail to see how someone who paid
those taxes isn't funding that research.

> I'm assuming that the additional research is small compared to the
> original research.

You may be making that assumption, but you're using a license that
applies the same restrictions to substantial rewritings, and even entire
programs that use the GPL'd software as only a very small part. Further,
forcing entire programs to be GPL'd even when they use just one small
GPL'd library is the stated aim of the GPL; see the whole story of
readline for clear example of that.

> If the additional research is the larger part, there's a chance that
> the corporation would have to rewrite the original software (interface issues
> etc). I agree with you though that in this case a BSD-like license would make
> more sense. But again, you don't know that when you're releasing the software.

Again, often you do. Do you think it wasn't obvious that the readline
library would under most circumstances be a very small part of any
program using it?

> My example was probably bad, but I don't see your point here. If all
> software used in the (actually more than) $100 laptop covered by
> BSD-like licenses had to be paid for, how much more expensive do you
> think it would be? I have no idea, but I would be surprised if it was
> more than a small fraction of the price.

Well, how much would it cost to independently develop an entire OS? Add
in an appropriate factor for the risk (since nobody's going to give you
the money to develop that OS for free), divide that by the number of
units, and there you go.

cjs
-- 
Curt Sampson       <cjs@example.com>        +81 90 7737 2974
           Functional programming in all senses of the word:
                   http://www.starling-software.com