[KLUG Advocacy] ESR: the GPL is holding back open source

[Robert G. Brown]

On Fri, 01 Jul 2005 21:48:40 -0400, Adam Tauno Williams <awilliam at whitemice.org> wrote:

>...I haven't seen much evidence that anything has really been
>stopped due to GPL induced fear.  Most of the time there are pretty
>clearly other causes of project failure or abandonment.

I've had to dispel a number of myths, primarily:

   Developing (or running) on Linux means that we must release everything
   we do to the whole world. 

>I think ESR is right in one regard - forking has not manifested as a
>significant threat, and the forks that have taken place (Samba-TNG comes
>to mind) have almost always fizzled out and died.  ESR talks about this
>quite a bit in the C&B and my perception is this is primarily what he is
>talking about.
Yes, that's true, and to a great degree, the framework provided by the GPL
(perhaps ANY licensing structure, but the GPL has been very effective)
tends to discourage that. IMO demolishing or deprecating the GPL without
replacing it with something else removes that framework.

>Taking code into a proprietary product also has proved,
>I think, to not be worth the effort;  for most solutions (except for
>HUGE hardware and software shops) it is really easier to just play
>along....
As a practical matter, it's very hard to adopt a part (a non-trivial
part) of a package without buying into the whole thing.

>...very willing to work with Open Source,  this is VERY different than
>ten years ago.
Nothing succeeds like success. 10 years ago, Linux was considered a toy
by anyone doing "serious" IT. Now that is has the imprimatur of some
organizations with some gravitas it's no toy, it doesn't look quite so
risky. 

>They may not agree with OSS on ideological terms... 
They don't understand it, but they have seen that IBM likes it, Hollywood
uses it to make movies, and it runs a lot of things at the NYSE. The 
rest is some philosophical mumbo-jumbo they don't have time to absorb,
but they got a memo from the Legal department that says it's OK.

>but the pragmatic benefits are undeniable (Do YOU want to create another 
>Apache?
Right.. of course not, that's even more risky than USING it! :)

>'faith in the free market' I assume refers to this quotation "Stronger
>virality punishes defectors more effectively, but also has more tendency
>to scare people away from joining the open source community in the first
>place. ..."
Well, yeah, to the extent that a lot of things in this interview are 
connected to a lot of other things. 

>"...Where the optimum point is all depends on how important punishing
>defectors really is relative to the economic pressures in favor of open
>source. My current belief is that the free market will do quite a good
>job of punishing defectors on its own; thus, increasing virality is a
>bad move."

But the free market has to be good at more then merely punishing defectors.
One might say (look at all the good spin-offs from large companies, many
of them are large companies themselves) that the free market, with the
incentives of ownership, ENCOURAGES defection.

I was thinking of other spots in the interview; his whole attitude
seems to be one of placing faith in a free market that doesn't actually
exist.

>I think he is correct,  and that increasing the virality <sp?> is a
>pretty radical move.  But I'm not convinced that a reduction in
>"virality" is in order. 
I think this is a side issue. The GPL isn't primarily about virality, but 
about protecting everyone's rights to see, share, and modify the source
code.

>The GPL seems a pretty fair balance to me, although what constitutes 
>"linking" seems to be growing fuzzier and fuzzier due... what constitutes 
>the delineation between one application and another?
I don't think that's too important. We don't really put "applications"
under the GPL, but source code. Now, we may CLAIM that the code is
intended, or suitable, for a particular application, but it's really
"just" the code that is covered. The "application" is a label. Some
may think this is splitting hairs, but it is really a very important 
aspect of this business, because (as a result of the GPL), we can take
some code and rework it into something the original authors did not
intend or so label their code as doing. 

As you rightly point out, "applications" are getting more intertwined
across networks, environments, platforms, and applications are becoming
fragmented. If you think of applications being GPL'ed, I see your 
problem, but if we stick to the code, the granularity doesn't change 
a thing.

A real monkey wrench could be introduced by writing licenses that erect
barriers to interoperability -- and that's not viral, but lethal.

><aside>The concept that the market needed to be persuaded by "viral"
>licensing, etc... to become "free" I find very telling of the
>fundamental flaw in the 'virtuous free market' that makes the entire
>concept untenable.  But I won't even get into that.</aside>

1. Many of the fears of free licensing are essentially viral ones, and
   the must be debunked before the suits can feel comfortable with these
   delivery methods.

2. ESR was instrumental in developing these arguments, and on the whole
   they have proved effective (enough so that if people want to go with 
   F/OSS, they won't let these issues stop them).

3. It's all very nice to speak of the "virtuous free market", but as 
   long as there are organizations that successfully strive to tilt 
   and warp the market in their favor with a welter of tactics unrelated
   to merit, quality, and performance, we don't have one.

>...I think in the full interview that ESR comes out as more moderate 
>than the captioned quotes might imply.
Yes,and a lot of it is focused on some odd topics, which are not really the
main point of what F/OSS is about, like virality and political correctness
and solidarity.

>He isn't talking about ditching the GPL for the LINUX kernel or
>pressuring developers not to select it.
I'm not so sure.

><aside>It is hard to discuss positive and concrete solutions with people
>who ideologically believe that 'the $SYSTEM will just take care of
>itself by itself'.  That is just something I believe to patently false...
I tend to agree. This is not a question of simply believing in some sys-
tem, but in defining values and creating a licensing framework that reflects
them.

>and I struggle to understand how anyone could ever honestly come to such
>a position.   I'm not stating that ESR categorically believes this,  but
>he seems to hold it as more-true than I do.</aside>

I can't tell if he is doing this as an exercise in discomfiting the F/OSS
community (which may be necessary), or because of some shift in his system
of beliefs or how to realize them. ESR certainly relishes the role of making
people uncomfortable, and then justifies it by telling us that the discom-
fort he brings is somehow good for us, because it shows that he is forcing
us to think outside the box.

I don't know what or where the box is, thus I don't know or care if I'm
thinking inside it or not. I don't feel wedded or committed to the GPL,
and I'm under no obligation to defend it. However, it does solve some 
problems in elegant ways, and as a tool of social engineering it has 
served its purpose well. I'd like to see ESR clarify what he feels ought
to come next and why, and lets hope it is based on todays conditions and
challenges. He is coming here (7 PM, August 9th, see our schedule!), and 
it is a good opportunity to question him on this issue.

							Regards,
							---> RGB <---