Re: [fluid-dev] Another application using FluidSynth announced

[David Henningsson]

On 09/07/2011 08:08 PM, Pedro Lopez-Cabanillas wrote:
> On Wednesday 07 September 2011, Matt Giuca wrote:
> [...]
> > http://lwn.net/Articles/396535/
> > In this situation, it was the Wesnoth team themselves that published
> > the game in the App Store (for a fee as well). Apparently the core
> > team were okay with it, but one of the contributors, Rusty Russell,
> > disagreed and raised a stink.
> >
> > He has a point. What good is a license if some (but not all) of the
> > team can decide to violate it later on?
>
> The game was released under the GPL. I don't see a license violation committed 
> by the release team.

Hmm, but whoever put it on the app store, or intended it to be put 
there, are either doing the violation themselves or actively promoting 
someone else (Apple) to do the violation.

> [...]
> > I will assume
> > we're dealing with the more blurry issue of people who release the
> > modifications' source code, but which end users technically can't make
> > use of (because the only way to install the software is to get the
> > unmodified version from Apple).
>
> You are assuming too much here. We don't know under which license Rouet 
> Production is going to release his product in October.

That is a fair point, and we should probably give them a ping before 
speculating too much, but looking from the front page of Rouet 
Production, it seems already available.

> > In that case, I would still say it is a violation of the *spirit* of
> > the LGPL because it violates this nice symmetry of the FS developers
> > saying to the downstream developers (Slide Control, for example), "You
> > can make modifications to this software however you like, but when you
> > pass it on, you must extend the same freedom to your users."
>
> And how do you know that they are modifying FS, and not respecting the license 
> terms?

We also don't know which version of FS they use (maybe 1.0.9, the last 
one not requiring glib?), so we don't even know who of us are copyright 
holders.

> > I can't say whether it violates the letter of the LGPL (v2), since I'm
> > not a lawyer. But my reading of it (as I went into detail on here
> > http://lists.gnu.org/archive/html/fluid-dev/2010-09/msg00028.html)
>
> Maybe you know much more than I, then. Is Paul Brossier, the guy that posted 
> that question, working for Rouet Production? Are them related somehow?
>
> > , is
> > that it's probably a violation of both Section 6 and 10. Section 6
> > says that you must "Accompany the work with ... the complete
> > machine-readable "work that uses the Library", as object code and/or
> > source code, so that the user can modify the Library and then relink
> > to produce a modified executable containing the modified Library" --
> > aside from requiring relinking, this also has the problem that most
> > users (those without an Apple developer license) will not have the
> > ability to produce a modified executable, so you cannot satisfy this
> > clause.
>
> You need a compiler and some other tools to produce an executable in any platform. To 
> compile FluidSynth for  Linux you need to accept a license of the GCC compiler as well, 
> which is also the official iOS and Mac OSX compiler, distributed by Apple under the GPL, 
> of course. You can download the Xcode package from Apple (containing GCC and other tools) 
> to build Mac and iOS applications, and it doesn't cost money. Note that 
> "gratis" is not required by the GPL, anyway.

I'm unfamiliar with exactly how development for iPhone works here. If I 
develop for iPhone, how do I put my own software on there? I mean, even 
Apple would think there should be a way to test your software on the 
real thing before publishing? Is that a legitimate path for distributing 
source code to e g FluidSynth?

> > Section 10 says that "You may not impose any further
> > restrictions on the recipients' exercise of the rights granted
> > herein," and the App Store does impose additional restrictions on the
> > use of the software.
>
> And _this_is_the_real_problem_ with Apple's App Store. By distributing a GPL 
> program with additional restrictions, they are violating the GPL. Not the 
> program authors, or the release team, but Apple.

IMO, since everybody know that this is what Apple does, I'd say whoever 
puts (L)GPL binaries there, is also responsible. Like giving out free 
guns to serial killers.

> When the VLC developers (http://www.videolan.org) protested against Apple 
> because the distribution of VLC in the App Store was imposing restrictions on 
> top of the GPL, the reaction from Apple was to pull VLC from the store [1], 
> because Apple don't want to respect the rights given by the GPL to their 
> customers, and of course they don't want to face a copyright lawsuit either. 
> The same happened with GNU Go [2].
>
> So, my opinion is that if you publish a GPL program in the Apple Store, you may 
> be losing your time because Apple will remove your product from the store as 
> soon as anybody raises his voice. There are other distribution channels for Mac 
> and iOS, anyway.
>
> And of course any product including FluidSynth, released for iOS or any other 
> platform, must respect the LGPL license of FluidSynth, choosing a compatible 
> license [3].

As for my own opinion, I tend to be pragmatic in the sense that I look 
for practical possibilities rather than the letter of the law.

But even if Rouet would plan not to give out source code to the extent 
needed to relink FluidSynth, one thing keeps bothering me. They actively 
chose to credit us by having our logo at startup, mentioning it on 
Youtube etc. This is positive, and they could have just have hidden that 
fact away and the chances anyone would notice it would be fairly small I 
guess. I'm just afraid that going after them with threats will lead to 
the next iPhone app developer just using FluidSynth without telling 
anyone instead, and likely getting away with it.

// David