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

[Pedro Lopez-Cabanillas]

On Wednesday 07 September 2011, David Henningsson wrote:
> 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.
 
Yes, there is moral responsibility in doing that questionable action. But that 
is something different to a license violation, that has legal consequences in 
addition to a moral reproach. 

I suppose that all FluidSynth contributors agree about the legal terms of the 
free software license used by our project, but it is very unlikely that we 
share exactly the same moral values, and that is not required either.  

Gnome is evil. There, I've said it.

> > [...]
> >> 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.

They have released several products, one of them is called "Slide Control". The 
product page says that "Slide Control is an Midi controller. NO SOUND is 
generate by the ipad. You must have a virtual synth or hardware midi instrument 
for use Slide Control." 
(http://itunes.apple.com/br/app/slide-control/id434680351)

The new product announced in the videos is a "Slide Control FluidSynth" bundle, 
promised for release in October.

> >> 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?

I can't give you answers, I don't have experience with it. I'm planning to 
start learning in a few weeks, though. If you want info about the development 
and testing process for Nokia devices, I know the matter :)

> >> 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.

That is again the semantic distinction between crime and sin. Well, indeed 
giving free guns to serial killers is already a crime. On the other hand, the 
GPL doesn't make any legal statement about helping evil companies to make 
bloody money.

> > 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.

I agree.

Regards,
Pedro