Re: [gNewSense-users] gNewsense: CDRTOOLS- gNS bug 108

[Kevin Dean]

Joerg,

Firstly, thank you for chiming in here. There is no doubt in my mind
that you have made many contributions to the GNU/Linux operating
system and for that I thank you.

On Nov 11, 2007 7:39 AM, Joerg Schilling
<address@hidden> wrote:
> "Kevin Dean" <address@hidden> wrote:
>
> > I'll disclaim this: I'm both a Debian user AND an Associate Member of the 
> > FSF...
> >
> > I've always taken Debian Legal's opinion VERY seriously.
> >
> > That said, there's something intestesting here beyond "do we accept
> > this license"... The question that keeps bouncing in MY mind is "Are
> > comments in GPL source code equal to license restrictions?"
>
> A few outsiders from Debian started a major FUD campaign against cdrtools.
> They came up with swashbuckling claims but never send any prove for their
> claims.

Ubuntu Dapper ships a version of cdrtools source that corresponds to
version 2.01, as such, gNewSense also contains this same version of
cdrtools.

In that source code, there is a file (libscg/scsi-linux-ata.c) that
says, and I quote directly:

 *      Warning: you may change this source, but if you do that
 *      you need to change the _scg_version and _scg_auth* string below.
 *      You may not return "schily" for an SCG_AUTHOR request anymore.
 *      Choose your name instead of "schily" and make clear that the version
 *      string is related to a modified source.

This is a very direct limitation on what a user can do with the source
code. I am not speaking of Debian's "FUD Campaign" as you call it -
gNewSense evaluates every software package regardless of what Debian
Legal or even the FSF says. gNewSense replaced Mozilla's Firefox with
BurningDog for one very simple reason: we can not ensure that everyone
who downloads the source from us has the right to release modified
versions of that software. This issue is irrelevant to the arguements
brought up before. They centered on license terms rather than the
practical "Can you take it, edit it and distribute it without limit."
In this case, someone who makes a change to this source must then
change what SCG_AUTHOR returns because having it return "schily" is
(as I understand it) a violation of a legal trademark you have
registered in Germany. I think it is a rather fair request - you
should not have to deal with harm to your "brand" for forked versions
of the software, it's essentially the same claim put forth by Mozilla.
It's a claim that gNewSense has a history of taking seriously, and
excluding or replacing software that has such trademark issues.

We're not accusing you of violating a license... There are binary
blobs in the Linux kernel that are licensed under GNU GPL that do not
meet our requirements despite their full compliance with the GPL.

>
> Distributions that asked lawyers publish the original cdrtools,
> distributions that switched to the broken ans questionable cdrlit
> trust in the "expertise" of laymen.

Most distributions also ship Mozilla Firefox and stay within the terms
of the MPL and Mozilla's trademark policy. We feel that since a user
can not download the Firefox source, make a change and then distribute
the changed version it does NOT meet the essential freedom criteria
required by gNewSense. To say "you may not call it Firefox" dictates
that the source can NOT be changed to call the application "Firefox".
Similarly, it is not possible to modify ONLY one part of cdrtools and
redistribute that program because

>
>
>
> > Thing on this: Shilly places a comment in the code that says "You
> > can't have it return Shilly"... if the software is GPL, one has the
> > right to remove that comment and one ALSO has the right to have that
> > function return "Shilly".
>
> You completely missunderstangthe GPL! The GPL allows you to use/modify/publish
> the code but it definitely does not cover trademark rights.

I understand that very clearly. I'm not arguing that you violate the
GPL by enforcing your trademark on cdrtools.

In this SPECIFIC instance, I was commenting on a "hypothetical"
issue... Technically, your warning about what SCG_AUTHOR returns is a
comment, as is the statement that "This is a hack." Beyond that you
declare your copyright and then indicate that the software is licensed
under the GNU GPL version 2 or later. The GPL specifically states that
other than the copyright notice (which is ONE LINE - Copyright (c)
2002 J. Schilling) and the license itself, you can freely edit and
modify the code found within - comments are code. Theoretically one is
as free to remove your Warning as your comment declaring "this is a
hack" - your warning (while understood to be a trademark claim) is
neither license NOR copyright notice.

THAT is what I wanted to point out. I'm not sure if you were notified
of this mailing when cdrtools became an issue or if you track the
gNewSense mailing list, but PRIOR to my message Paul O'Malley brought
up comments to the extent of "lets think on licensing for a moment".

>
> Thr protection is unfortunately needed to grant the _users_ a working
> cdrecord under the name "cdrecord".

While I understand your motivation to protect the "brand" and ensure
that cdrtools is high quality, I feel that by licensing and releasing
cdrtools source as Free Software you give up your right to control how
it is used. I stand by your right to limit the use of the cdrtools
name for non-official version - I stand against your right to dictate
what a function in a Free Software application can return.

>
> If there were no Linux distrubutions that deliberately add bugs
> to the code, there was no need for this extra protection...

gNewSense is a rather small distribution with a small developer base.
Our preferred method to "fix" this issues would be to re-brand
cdrtools to comply with your trademark enforcement AND to release high
quality Free Software that protects people's right to share useful
tools. However, since your trademark also affects the output of that
application it is impossible to make a perfect "drop-in" replacement
of the software that both meets our criteria and your trademark
rights. Since we don't have the ability to fully anticipate how this
may affect other applications in our repository or our abilty to
managed those changes, our most practical solution is to track the
work of another distro, in this case Debian comes closest.

> Now the "trademark" cdrecord is something that disallows bastardization
> to the detriment of the users.
>
> If you like this to go away, talk to the Linux distributions that
> deliberately added bugs: Suse RedHat and Debian.

With all due respect, you mention the "distros that talked to
lawyers", I'm quite sure both Red Hat Inc, and Novell have quite a bit
of lawyer interaction considering the fact that they're the putting
multiple millions of dollars on the line by not complying with license
and trademark terms. I think it speaks directly to the issue that to
ensure their freedoms to edit the code they needed to "bastardize" the
application.

>
>
> Jörg
>
> --
>  EMail:address@hidden (home) Jörg Schilling D-13353 Berlin
>        address@hidden                (uni)
>        address@hidden     (work) Blog: http://schily.blogspot.com/
>  URL:  http://cdrecord.berlios.de/old/private/ ftp://ftp.berlios.de/pub/schily
>