Re: Economics and the GPL

On Wed, 12 Jun 2002, Matt Seitz wrote:

> Probably.  What if the GPL itself had such a stipulation in its own license. 
> In other words, the FSF could change the license on the GPL itself from
> "Everyone is permitted to copy and distribute verbatim copies of this license
> document, but changing it is not allowed." to some terms that said that the GPL
> could only be copied and distributed with works that are exclusively licensed
> under the GPL. If the work was distributed under the GPL and another license,
> then the FSF could sue for breach of the license for using the GPL.

I don't think so.  They might be able to sue for copyright infringement if 
you modify the GPL without authorization.  But they can't stop you from 
using their GPL verbatim, whatever the terms, and then blatently violating 
those terms.  After all, as the copyright holder, you're the one with the 
standing to enforce those terms, not the FSF, and you're not likely to sue 
yourself for compliance, are you?

> > > Even RMS accepts that placing restrictions on unauthorized mass, commercial
> > > duplication may be necessary to make funding of works possible.
> > 
> > And does he suggest that can be done in a way compatible with the GPL?
> 
> No, what he suggests [in his writings and speeches] is that if copyright laws
> were changed to allow personal copying and the other freedoms he advocates,
> then the GPL would no longer be necessary.  In other words, the GPL is more a
> protest tool than the ultimate goal.  He sees the current copyright system as a
> weapon of oppression, and seeks to change the system by turning that weapon
> back upon the oppressors.   

I'm not sure that unlimited personal copying is necessarily compatible with 
the public policy goals of "the copyright bargain".  Perhaps there's some 
way to allow very limited amounts under "fair use", but that would take 
some effort to figure out just how much is fair and how to keep it from 
getting out of control.  Remember, Napster looms large in the minds of many 
content producers.  (Nevermind there's no proof it really harmed CD sales.)

It's also unlikely for copyright law to be modified to demand source like 
he wants, so I don't see that happening either.

The one reasonable goal he could hope for is a reduction of copyright term 
to a more reasonable duration.  However, this is an uphill battle against 
powerful entrenched interests, so it's hard to win this one.  Still, this 
is arguing in the public interest, and librarians, legal scholars and many 
others are fighting this battle already, not just "commie whackos".  ;-)

> > If the GPL had the code "expire" and fall into the public domain or 
> > BSD-style free licensing after a fixed time period (e.g. 5-10 years), I 
> > wouldn't worry as much...
> 
> I think that would be a great improvement to the GPL.

Good luck getting Stallman to go for it, though.

> > Of course, Wine has sort of done the reverse -- it was well-established 
> > under the BSD/X11 licenses, and switches to LGPL.  
> 
> netatalk recently made a similar switch from BSD to GPL.  My impression was
> that it was more because of a theoretical concern about code stealing than
> because of any specific event.

Is this a trend?  Maybe due to the influence of GPL advocates, in part?

> > The BSD camp makes an interesting argument, but I can't fathom how to 
> > construct a valid test to figure out if they're right.  There's too many 
> > other factors you wouldn't be able to control, so it's all anecdotal...
> 
> I was thinking about that, too, and came to the same conclusion.  

Still, it would be nice to test empirically, somehow.  But how?  Hrmm...

> > If anyone could use them, Microsoft would be.  
> 
> Microsoft can.  In fact, Microsoft does.  Microsoft distributes various GPL
> tools as part of their Microsoft Services for Unix product (see
> http://www.microsoft.com/windows/sfu/default.asp).

Do they?  Interesting.  Well, that probably falls under the category of 
"mere aggregation", and if they include the source, that should be 
copasetic, even if it's the evil empire doing it...

> > Now, it's true that membership is open to the public, but under 
> > conditions that are unacceptable to some.
> 
> Now we get into questions of how free is free.  Even public parks have 
> rules on conduct.  If one finds those condtions unacceptable, one will be 
> asked to leave the park.  Does that make them no longer "free" or "public"?

People might get into trouble if they're being disorderly and stopping 
others from being able to enjoy the park, but they're not asked to sign a 
contract to behave upon entering either...

> > It ["community"] doesn't carry the '60s hippie connotations that 
> > "communal" does.  
> 
> Exactly. It's funny how chaning "al" to "ity" can change one's perception.  

Indeed!

> > I've been considering trying to create a community that doesn't 
> > necessarily support "free beer" code, but otherwise could have the
> > sorts of benefits the GPL community derives from the GPL.  
> > 
> > So, do you think it could work?  
> 
> I think it's worth trying!

Sun's attempt with SCSL wasn't very well-received.  Then again, Sun is a 
for-profit company with a reputation for being control freaks, so that 
might have something to do with it.  Still, SCSL might color people's 
perceptions of the next "community source" or "community software" type of 
community someone tries to create.  Then again, they've laid some of the 
groundwork, which might help gain some acceptance.  I don't know.

> > How could I avoid backlash from free software/OSS ideologues who would just 
> > assume it must be bad just because it doesn't conform to the FSG/OSD?
> 
> Well, for OSS folks, the primary goal is to have source code available to the
> users.  So as long as you gave the members source, and allowed members to
> distribute fixes to other members, I think you could get some support.

Well, that's what I was hoping.  Being of a largely OSS mindset myself, my 
interests are pretty similar.

> For FSF folks, their two main goals are to have source code available to 
> the users and to allow personal redistribution.  For the first, see 
> above.

This part should be okay, I think.

> If you also allowed members some limited individual rights to 
> redistribute code to nonmembers, you might reduce much of the FSF camp's 
> opposition.  Yes, there would still be a "free beer" aspect.  But the 
> idea is to allow giving away six packs, but prevent building a brewery.

This is a good point.  I suppose, like the possibility of semi-proprietary 
commercialization within the community (e.g. paid subscriptions to a given 
project), this sort of thing could be optional, at the discretion of the 
project.  I can see two options for potentially allowing limited personal 
redistribution which might be reasonable here.

First might be to allow redistribution and use, but not allow the code to 
be received from another nonmember.  That is, members of the community 
might be permitted to distribute the code to nonmembers, but a nonmember 
would have no redistribution rights at all, possibly including their own 
modifications, even.  (There should be some incentive to join.)  This would 
be the more liberal option, but it might be too easy to get the code for 
free, largely defeating the purpose of restricting it to the community.

The more restrictive option is to allow such redistribution only on a 
temporary basis for "evaluation" purposes only, like shareware.  Sure, some 
people will ignore this restriction, but at least then they'd be violating 
the license and could be asked to join the community to become compliant.
This approach is probably the most reasonable one to officially endorse, 
but even this one should be optional -- some project authors may not want 
to be this lenient with nonmembers of the community.

Of course, unofficially, the community could simply recognize that there 
would likely be a certain amount of "piracy" where unauthorized copies are 
distributed to nonmembers.  As long as any crackdowns on this aren't really 
draconian in nature, this could serve as a natural marketing mechanism, as 
"piracy" (unauthorized copies) benefits Microsoft.  Their monopoly might be 
less complete if EVERY copy was really properly paid.  There are a lot of 
users who use pirated Microsoft products because they're easy to copy, and 
those people are still helping the network effects to get more people to 
buy legitimate copies as well...

So perhaps a small, restricted amount of personal copying could be allowed 
for some projects, but it would be up to each project, and it would have to 
be limited enough that users should prefer to join the community.

Deven