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

Re: distributing precompiled binaries

On Thu, 2 Apr 2009 18:52:45 +0200
Francesco Poli <frx@firenze.linux.it> wrote:

> On Wed, 1 Apr 2009 23:02:06 -0700 Steve Langasek wrote:
> [...]
> > It's reasonable for you to hold the position that this is "not
> > free".  But that's not what the DFSG says; and before someone tries
> > to change the DFSG to say this, I would recommend someone try to
> > come up with a brighter line to separate documentation than, e.g.,
> > fonts, graphics, sounds, and videos than just "more people edit
> > documents".
> As far as I am concerned, I do *not* want to separate documentation
> and programs from fonts, graphics, sounds, and so forth.
> I am convinced that *all* these works need to have source available in
> order to comply with the DFSG and be called Free.
> And I respectfully disagree with your claim that the DFSG don't apply
> (or apply in a weaker sense) to documentation, or graphics, or sounds,
> or whatever.
> As Bruce Perens clarified, the original intent was for the DFSG to
> apply to everything distributed in Debian (main):
> | When designing the DFSG, I was considering the contents of a Debian
> CD, | much like the Official Debian ISO image, containing all of the
> Debian | software and documentation.
> [...]
> | I intended for the entire contents of that CD to be under the rights
> | stated in the DSFG - be they software, documentation, or data.
> Please see
> http://lists.debian.org/debian-legal/2003/08/msg00264.html
> for the context.
With all due respect, I think you are mixing issues here. I don't think
anyone has advanced an argument that anything, including data, should
not be covered by the DFSG. The question is, for those types of data
that are not reducible to text files, what considerations should govern
the format(s) in which such data is provided. 

If an upstream author chooses to offer non-text data under a free
license, the format in which the data is offered should, in the first
instance, be within the discretion of the author. Artwork, music, and
video can be supplied in a variety of formats or containers. Accessing
and modifying that data in a useful way may require specialized tools
and knowledge. Unless the required tools are non-free, none of this
makes the data non-free.

Similarly, if a hypothetical upstream author creates documentation that
combines text, images, and formatting (or formatting alone) and then
offers that documentation under a free license, that is in fact a
desirable result. Maybe that documentation takes the form of a mindmap
file, or odf or pdf, or something else. In any of those cases, the
format does not make the data non-free. With the right tools and
knowledge, the data in any of these formats can be accessed and
modified. This is in contrast to a binary blob of compiled code, which
can be only approximated by disassembly or reverse engineering. (And
this is the crucial distinction.)

Now there may be any number of reasons why some formats might be
preferred and encouraged, and other formats deprecated and discouraged.
Ease of accessibility and modification (which may not be the same
thing) are, without doubt, factors to consider. That debate can be had,
and decisions reached, without resort to labeling an author's choice of
data format as free or non-free.

Reply to: