[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

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

From: David Henningsson
Subject: Re: [fluid-dev] Another application using FluidSynth announced
Date: Tue, 13 Sep 2011 09:58:56 +0200
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:7.0) Gecko/20110906 Thunderbird/7.0

On 09/12/2011 11:38 PM, Pedro Lopez-Cabanillas wrote:
On Monday 12 September 2011, David Henningsson wrote:
On 09/11/2011 09:28 PM, Pedro Lopez-Cabanillas wrote:
On Sunday 11 September 2011, David Henningsson wrote:
On 09/07/2011 10:38 PM, David Henningsson wrote:
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?

     From what I can understand the development tools are free to download
and use, but testing your software on the iPhone or iPad costs $99 per
year. [1] Interestingly enough, this is a relatively low fee compared to
buying an iPhone/iPad.

This is true, latest Xcode release is available for download in the Mac
Store at no charge:

Older releases where in the second DVD, named "Mac OS X Install Disc 2"
distributed with the Macs, like mine. I've downloaded also some version
Apple as well, and I'm not a member of the iOS Developer Program.

But as I've said, if the compiler and developer tools are "freeware" or
not is
irrelevant from the license point of view, in my opinion.

Do you consider the $99 fee part of the compiler in this case?

No. It is the annual membership fee of a private club. This is how Apple
wanted to organize his community, around the maximum values of greed and
selfishness. They have the right to do so, no matter how unethical this looks
to people with a very different set of values. Don't be fooled by things like
this: http://www.opensource.apple.com , they are evil.

Because this ethical distance, it seems to be unfair that their broad
community, including all their users, can enjoy our work. And we may find some
legal argument to deny them that right, even if to do so we need to bend a
little or force the letter of our legal documents, because it is clear in our
soul what is the true spirit. [This was a bit sarcastic, of course; more

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.

Let me clarify this a bit. LGPL contains a lot of rules and regulations,
but let me point out the two types of freedom, that the end user is
given, and that are important to me:

1) Available source code. I e, if the developer fixes a bug in
FluidSynth and makes that version of FluidSynth publicly available, we
should be able to take that fix and incorporate it into the next
FluidSynth release, and release that version under LGPL.

In this case, Rouet can fulfil that requirement by publishing their
FluidSynth source code changes, or publicly state that they haven't done
any changes.

2) Updating FluidSynth. If the end user finds a bug in FluidSynth by
playing around with "Slide control", he/she should be able to fix it and
run the fixed version on the same hardware. The question is if Apple
fulfils its part of this deal by the $99 developer program [1]. On one
hand, the program is widely available and relatively cheap, on the other
hand, it's still a cost, and Apple can probably choose to deny this
program on an individual basis. So I'm not really sure what to think
about it at this point.

Anyway, Rouet can fulfil that requirement by publishing the entire
source code to "Slide control". I don't know if they can also choose to
supply some kind of linkable object code, that depends on what you can
and can't do with XCode.

No objection. I think that documenting a clarification about this
interpretation of the license would be enough,not requiring a license
addition explicitly accepting exceptions, or license changes.

Note that the above is not a full interpretation of the license, I'm
just pointing out what is important to me. There might be other problems
with the App Store distribution channel, which are important to someone
else of FluidSynth's copyright holders. (Especially taking into account
the code already present when I got involved with the project.)

Likewise, for me it is not only important the letter of the LGPL license, but
the ethical principles inspiring the libre software movement. For instance,
the principle of no discrimination that is not part of the LGPL, but is
included in the Open Source Definition, the Debian Social Contract ("No
Discrimination Against Fields of Endeavor") and the free software definition:
"freedom to run the program, for any purpose (freedom 0)".

Forbidding FluidSynth in the App Store means to deny all the iPxxx users even
the knowledge that this project exists, not to mention the right to use the
program. In the long run, this means more harm than good for us.

Finally, I'm absolutely against the feeling of moral superiority, of
fanaticism and radicalism.

There are two dimensions here. First, there is the issue of whether we can take action against App Store, and that depends on how we interpret the LGPL, and second, whether we actually proceed with taking that action, which depends on our spirit, soul, moral, and all that.

For the first dimension, it is not obvious to me whether we can or not, but it seems likely, especially if we, as you say, "bend a little or force the letter". So the precondition for taking the app down seems to me to be fulfilled.

You're right in that many of us dislike Apple's attempts to lock the platform (myself included), and that it is important not to let that cloud our judgements when we need to decide what's best for the project. But regardless of what we think of Apple, and if bringing that app down means harm or good to the project, it just takes one of all copyright holders to raise a complaint to bring the app down. That includes all copyright holders in the past which we know nothing about.

This actually comes down to another question. Does the project need to protect itself from that scenario? If so, we need to relicense FluidSynth, e g under BSD or under GPL with Classpath exception. That is done by asking the contributors we can get hold of to relicense, and rewriting the code for people that refuse or that we can't get in contact with.

So which is worse? Relicensing and rewriting parts of FluidSynth, or denying FluidSynth for iPhone/iPad users?

If you ask me, I will personally not commit to doing the job of contacting copyright holders and ripping out code that does not fulfil the new decided license. I will, however, prefer to relicense my own contributions if the option is to have my own code ripped out.

// David

reply via email to

[Prev in Thread] Current Thread [Next in Thread]