[LAU] Aeolus

Rob Kudla lau at kudla.org
Wed Sep 18 23:44:58 UTC 2013

On 2013-09-18 17:16, Fons Adriaensen wrote:
> Neither of them even had the decency to let me know of their work, and 
> both are taking Aeolus in a direction I do not approve of. Gavioli has 
> even added his 'copyright' to the sources of the libraries that Aeolus 
> depends on but which are not part of its source distribution. Apparently
> the intention is to release incompatible versions of those as well.

If he's made changes to the libraries, he does own the copyright on his
changes -- but only his changes. Diff yours and his, and that's his work.
That doesn't affect your trunk, though, just his branch. And of course, if
he removed your copyright details, or they weren't in every file to begin
with, he's violated your license and therefore is infringing your
copyright. (I can't seem to find his code, so I have no idea.)

What he doesn't own is the "Aeolus" name, and I'd think a nice email would
straighten that out right quick. (I'm surprised you never sent one to Red
Hat over their recently discontinued Aeolus Project.)

But being able to fork is one of the huge strengths of the free software
model, as well as a weakness when people don't follow basic etiquette like
naming the fork something else. There may still be issues (c.f. the ffmpeg
vs. avconv hissyfit where the Debian maintainer preferred avconv and
included a program still called "ffmpeg" in the package, which prints a
"this program is deprecated, use avconv instead" message every time you run
it, there is no proper ffmpeg available except through PPAs, and its
command line options are just different enough to make ffmpeg wrappers
puke) but when you cut to the chase, people being able to make their own
versions of other people's software is the entire point of free software.

And if they do something you like, you can use it too. It's not a bug, it's
a feature. Surely you see the value of, say, Yoshimi, which began as a fork
of zynaddsubfx? (I certainly do, not least because I'll never again have to
see religious propaganda in the readme.)

If he altered or removed your copyright notice, though, that's not part of the
free software model, that's simply copyright infringement.

> And I will make sure that this sort of thing won't happen again, even if
> that means a more restrictive license.

Well, that's your right, of course, but ask the OSS guys how well that
worked out for them. If you change your license to a non-free one, one of
the forks might very well replace Aeolus in, for example, Debian and its
derivatives. I guess you could go proprietary, but then you're competing
with free and your name is not Apple.

As things stand, the version of Aeolus available to me as an Ubuntu user is
yours, not one of the forks as far as I can tell. I suspect the same is
true of other Linux audio friendly distributions. Free software is
(normally) a meritocracy. What would make a distro pick theirs over yours?

What direction are these other guys taking their forks in that bothers you,
anyway? I mean, it seems to me that they're probably making a built-in
instrument for use in their own software (like LMMS' embedded version of
Zyn), not a viable competitor. If their code is sloppy or buggy or limited
to one use case, no one's going to pay them much attention regardless. I'd
have never known they existed if you hadn't mentioned them, and as I said
above, I can't find their code.

This may even all be a misunderstanding, as I found a bug report on the
Musescore bug tracker from a year ago in which someone posted a link to
your old skynet.be page with Aeolus 0.6.6 as the latest version, which made
me think you'd let it languish till I thought to type "dpkg -l aeolus". If
what they're doing is a fork of 0.6.6, I'd say that's the most likely
situation, which would also explain why they didn't reach out to you --
especially if you don't get mail on your skynet.be email address anymore.


More information about the Linux-audio-user mailing list