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

Re: various opinions on Debian vs the GFDL

Scripsit Anthony Towns <aj@azure.humbug.org.au>
> On Thu, May 08, 2003 at 09:24:21AM +0200, Henning Makholm wrote:

> > We're not talking about music; we're talking about *sound
> > recordings*.=20

> Actually, we're just talking about embedding sound in a GNU FDL document.
> Music, in case you hadn't noticed, is one form sound takes.

But not the only one.

> > All the XML scores in the world will not allow me to
> > recreate a particular sound recording (made with real live musicians,
> > in the case it contains music). Therefore, an XML score is not
> > source.

> All the C code in the world won't let you recreate the last build I did
> either, unless I also give you the compiler I used. Big deal.

If one need to use a compiler that only you have, I'd say that your
binary is not free.

> > > Samples and recordings are more difficult, mainly because the concept of
> > > "revision" doesn't really exist, per se. One possibility is just to do
> > > a hex dump -- it's as straightforwardly editable with a hex editor as
> > > _anything_ is, after all

> > Any opaque format is straightforwardly editable with a hex editor.

> Well, no, it's not.

It was your claim that sound could be edited in hex form. If that is
true, then anything else can be similarly "edited" in hex form.

> The question is what changes do you want to make.

Nowhere in the GFDL does it say that it is OK for a transparent format
to make only certain kinds of changes possible.

> If you want to change the location of two icons in a program, I don't
> think you're going to be able to do that if I give you a hexdump of an
> ELF executable.

And if you want to change the words of a poem read alouf, I don't
think you're going to be able to do that if I give you a hexdump of
a PCM file.

> OTOH, I don't think there are any "revisions" you can make to any
> sound file that you can't also make with a text editor to a suitable
> text dump of a WAV file.

My point is exactly that *no* way of editing sound files will allow me
to do the kind of changes we normally require for freedom.

> > Only for certain kinds of changes. That's not enough.

> Really? How do you remove all the buffer overflows from mozilla with
> a text editor? A lot of analysis, study, and tedious editing, no?

Yes, but it's possible in principle.

> Same thing with most of the edits you want to do to a sound file.

No, they are not possible in principle.

> > Not without losing any semblance of sensible prosody.

> Again, so what?

So I cannot reasonably make trivial edits and have results that have
reasonable quality.

> The sorts of revisions you can do with sound are fundamentally
> limited;

Exactly. Therefore, sound is opaque no matter what format it is in.

> That's completely irrelevant too: the question that's answering is whether
> the formats specifically designed to thwart modifications.

The "thwart modifications" language in the GFDL applies only to
unconventional uses of otherwise transparent formats. The definition
of transparency is;

| A "Transparent" copy of the Document means a machine-readable copy,
| represented in a format whose specification is available to the
| general public, that is suitable for revising the document
| straightforwardly with generic text editors or (for images composed
| of pixels) generic paint programs or (for drawings) some widely
| available drawing editor, and that is suitable for input to text
| formatters or for automatic translation to a variety of formats
| suitable for input to text formatters.

I maintain that this effectively excludes any conceivable sound

> > If it is not possible to license sound under GFDL (which I believe it
> > is not), then the GFDL says that I must not make a modification of
> > the work that consists of reading it aloud on a sound recording. I
> > think that's quite easily non-free.

> That's wrong too: that would merely be an opaque copy which is entirely
> allowable, as long as you distribute a transparent copy as well.

I *cannot* distribute a transparent copy of my spoken performance,
because no such copy is possible, as argued above.

Henning Makholm                         "Al lykken er i ét ord: Overvægtig!"

Reply to: