Hello,
[at the end follows a short sumary for people who don't want to read it all]
First I answer Manoj's post, and further below I answer Jules.
On Sat, Aug 08, 1998 at 01:54:57AM -0500, Manoj Srivastava wrote:
> Hi,
>
> There are two nebuously related ideas in this message.
Yes, the primary goal is to seperate things now. I'm happy that we steer in
the right direction.
> I think I want to differentiate between the kinds of changes
> that we are talking about here. If I write a standards document, I do
> not want people subverting or otherwise modifying the contents (the
> wrods that I wrote) -- at all.
I think this requirement is sufficient to make it dfsg free. I would be
opposed to include it in main if it is not modificable at all, but a name
change is really accepted by everyone.
> If they do, I want them to call it
> something different. However, I could not care less if the converted
> it from postscript to text to pdf or rendered tiff, as lon as someone
> reading the document sees the same words in the same sequence.
>
> (modification means chaging, deleting, or adding to the words and
> images that make up the document, format conversion themselves do
> not constitute a modification)
Yes. Those two things are independent, and have to be granted in the
copyright explicitely. This is the reason why I added the changed GPL at the
end of the message, where those points are expressed in point 2
(modification) and point three (format conversion). In the GPL, point three
implies point 2, but this can be changed for people who don't like
modification but only foprmat conversion.
> Any license for documentation, and any policy that Debian
> institutes, should differentiate between these kinds of changes.
Absolutely.
> I personally think that it would be permissible to accept
> any document, including a standard, which is distributable with the
> following restrictions:
> a) the document is distributed unmodified along with patch files,
> b) The document is clearly marked as changed, and,
> c) the document has a different name.
Yes, this is my opinion, too.
In addition, it should allow format conversions of the original and the
modified version with the restriction that the converted document carries
the same copyright. Some people would not like this (the Gimp Manual
restricts this because they want a high quality print only), well, but I
think it is not good to require a name change when only format conversion is
done. Maybe it is okay to require that this is marked as an inofficial
conversion?
> ______________________________________________________________________
>
>
> However, there are things (like a magazine cover, or a
> graphical novel, where layout and formatting are an integral part of
> the document, and modifying or altering them would detrimentally
> affect the document/piece of art.
>
> As far as Debian is concerned, we should bear in mind that we
> could be looking at documents that go beyond mere software
> documentation, and I would like to see tham in main as well.
I would not call it "software documentation", but "technical documents", as
the latter covers more. The German word is "Sachtexte", but I can't find a
translation. Anybody has a better dictionary?
So, yes, we have no *interest* in changing the cover of a magazine, or a
graphical novel (which expresses opinions in some way), so this is covered
by the Exception section I posted along with the dfsg interpretation. I
really think we can talk about inclusion of non-free non-technical
documents in main (for example: a computer magazine could restrict changes
to the magazine cover, but not to the content of the articles, to remain
free enough. I think that this is "hitting the border line", and argueable,
but for me it's enough).
> So, if someone creates a graphic novel, that tells a story,
> and distributes it freely in pdf format; allowing no modification or
> conversions away from pdf; why do we need to change anything? Why
> would we try and modify it after the author is done? Why should this
> not be accepted in main? The modification clause may make sense for
> compute programs, but for the wider domain of documents, I think it
> may not make sense.
Yes, this is why I think we should make exceptions to the dfsg. The dfsg
would be the official definition, and in my text I wrote something about
possible
exceptions. Did you read the second part of the text I posted? There I had
several suggestions for exceptions: Emails, Essays (a graphic novel is would
belong to this group of data entities), personal opinions and copyright
licenses. Although I didn't named graphic novels explicitely, I think they
are covered there.
Those are not dfsg free, but we should allow the inclusion in main anyway,
because we don't benefit from changing the content.
Up to this point, I think you are completely right.
> I think we should relax the modification requirement for
> anything that happens not to be a software programs
> documentation. (even standards should be acceptable is they allow
> modification with name changes/ patches)
Here I would extend a bit from software programs documentation to technical
documents in general ("Sachtexte"). I consider standards to be technical
documents. But a name change and distribution of the original source along
with patches is really all we need, so they are already acceptable by the
dfsg definition, and we don't need exceptions for them.
On Sat, Aug 08, 1998 at 11:19:43AM +0100, Jules Bean wrote:
> On 8 Aug 1998, Manoj Srivastava wrote:
> >
> > I personally think that it would be permissible to accept
> > any document, including a standard, which is distributable with the
> > following restrictions:
> > a) the document is distributed unmodified along with patch files,
> > b) The document is clearly marked as changed, and,
> > c) the document has a different name.
> >
> > I see that marcus agreed to all these while discussing the
> > dfsg.
>
> Absolutely. As far as I am concerned, that qualifies as modifiable. As
> you will see from my proposal (the informal one which predated Marcus's by
> a day or two), I personally find any combination of those 3 requirements
> (as determined by the author of the original) to be acceptably 'free'.
It's great that this seems to be a general agreement. People who disagree
that this is the things the current dfsg mandates should speak up now.
> > However, there are things (like a magazine cover, or a
> > graphical novel, where layout and formatting are an integral part of
> > the document, and modifying or altering them would detrimentally
> > affect the document/piece of art.
> >
> > As far as Debian is concerned, we should bear in mind that we
> > could be looking at documents that go beyond mere software
> > documentation, and I would like to see tham in main as well.
> >
>
> I disagree. They're not free.
Yes, but the point is that they are no technical documents, and we don't
benefit from being able to change them. In some ways they express a personal
opinion (you read the RMS text, did you?). However, although I find the
magazine cover hard at the border line, I think a non-free novel is
acceptable. A novel never has a "bug", we can never "fix" it. It is correct
so as it is by definition (it's like history, you can't correct what
happened). Very unlike technical documents.
> > So, if someone creates a graphic novel, that tells a story,
> > and distributes it freely in pdf format; allowing no modification or
> > conversions away from pdf; why do we need to change anything? Why
> > would we try and modify it after the author is done? Why should this
> > not be accepted in main? The modification clause may make sense for
> > compute programs, but for the wider domain of documents, I think it
> > may not make sense.
>
> I personally believe that something like this is not free, and should not
> be in main. We are not in the business of distributing graphic novels,
> after all. Whilst I respect the wishes (and copyright) of any author, and
> I fully understand that they might not want to allow modification to their
> works, I do believe that the resulting work is non-free. So we can put it
> on our FTP site, but we should not put it in main.
We should make a distinction here. Would you also object licenses (which are
not modificable) in main? What about emails (for example the email from
Linus about the kernel-header policy). And what about personal opinions? For
example, an interview with RMS in the emacs package. Would you want to
change them?
We should always ask us "Why do we need it to be free?". There is a big need
for freeness in software, and also in documentation. But there is no need
for a free novel, alithough it would be nice to have one, the result may be
quite confusing :)
Please take a look at the "exceptions" part of my posting, and let me know
which of the exceptions you would allow and which not.
> > I think we should relax the modification requirement for
> > anything that happens not to be a software programs
> > documentation. (even standards should be acceptable is they allow
> > modification with name changes/ patches)
>
> I disagree. If it doesn't meet our criterions of free-ness, we do not put
> it in main. Most of the documents we are likely to distribute - manuals,
> HOWTOs, FAQs, standards - all benefit from being free. I have no problems
> with those documents which don't benefit from being free - original,
> copyright-enforced works of art - going into non-free.
Okay, I see now that you are well aware of the fact that we don't profit
from freeness of works art. But some non-freeness must be included in main
(at least some emails, the copyright licenses in /usr/doc/copyright/* and
/usr/doc/*/copyright, anmd some other files).
If we restrict ourself to technical documents, we loose personality,
humanity. We should allow the expression of personal opinions in essays
about free software, for example.
--------------------------------------------------------------------------
I'll try to summarize the two main issues:
a) Manoj wrote:
> I personally think that it would be permissible to accept
> any document, including a standard, which is distributable with the
> following restrictions:
> a) the document is distributed unmodified along with patch files,
> b) The document is clearly marked as changed, and,
> c) the document has a different name.
This is only one aspect, as we need other rights, too (format conversion,
distribution of the patched and format converted "binary", with and without
fee), but I think this is the core. Is there anybody who think that we can't
grant any of the three things above?
b) Non-technical documents.
I listed the following:
1. Trademarks, Copyrights
2. Personal opinions, email quotes
3. essays, graphic novels
I think everybody agrees that we don't benefit from changing these, and we
shouldn't change them. (As this would probably be a violation of the
personality of the author). Are they only acceptable in non-free (as Jules
says), or do we allow them in main (Manoj and me)?
Thank you for your input,
Marcus
--
"Rhubarb is no Egyptian god." Debian GNU/Linux finger brinkmd@
Marcus Brinkmann http://www.debian.org master.debian.org
[EMAIL PROTECTED] for public PGP Key
http://homepage.ruhr-uni-bochum.de/Marcus.Brinkmann/ PGP Key ID 36E7CD09
