DRAFT document on coding conventions in kde libraries

[Nicolas Goutte]

On Wednesday 08 March 2006 23:44, Lauri Watts wrote:
> On Monday 06 March 2006 12:41, Nicolas Goutte wrote:
> > On Sunday 05 March 2006 22:53, Friedrich W. H. Kossebau wrote:
> > > Am Sonntag, 5. März 2006 18:38, schrieb Olivier Goffart:
> > >
> > >
> > > Are pointers sufficient like "See File xy for authors, copyright and
> > > licence" or should there be a complete documentation of all authors
> > > and the
> > > licence (although they might be fetchable from the repository system)?
> >
> > Basic copyright and license information *must* be in the same file or
> > some "attacker" could (pseudo-)naively tell: "I did not know".
>
> This is a myth actually, albeit a widespread one.  Copyright is not
> license, and "I did not know" is no defence at all if it comes to the law. 
> Unless something specifically says it's public domain, or it has come into
> the public domain by virtue of being older than the prescribed number of
> years (and we're quite many years from any parts of KDE hitting that limit)
> , it is _always_ copyright someone, even if we don't always know who that
> someone is or how to reach them.

Sorry, I am not an expert, even less a lawyer but I do not think that it is a 
myth, perhaps only a difference between law systems. (You seem to refer only 
to the Berne convention and U.S.-specific stuff below.) And perhaps it is not
the law either but just how to be able to correctly prove in a court case 
(and that in any country) that one persion is the copyright holder, without 
massive cost (money or effort).

Also I have not meant that not having a copyright means that there are no 
copyright. Sure, every developer/artist/author/translator/... having 
committed a non-trivial commit in SVN has a copyright on that file, even if 
it is not described explicitely in that file.

As for not having the license in the file, that is surely more problematic, as 
we have no idea about the license of that exact file (e.g. many Perl/Python 
scripts in KDE). So such files having only the minimal license: "You can use 
it for KDE" which is not really open-source.

As for "I did not know", sorry, in a world when you can get millions when 
telling "I did not know that the coffee was hot", you can do many things with 
"I did not know" (even more if lawyers will get millions if they win).

>
> > > And what sense
> > > make the notion of the year span at all, when to update?
> >
> > Year spans (e.g. 1996-2004) are problematic, as far as I have understood,
> > and listing each year has to be prefered.
>
> The years are "year of first publication".  Each change, when distributed,
> therefore should be added (because newly added code is being 'published'
> for the first time.)
>
> So, you list each year a change was made, and may abbreviate that to a span
> if multiple consecutive years are included.
>
> e.g. File was change din 2000, 2001, 2002, 2006 (unlikely, but possible)
> You could put "2000-2002,2006"
>
> Most important of all, is to realise that just _creating the file_ grants
> you copyright (this is called 'The principle of "automatic " protection'). 
> The copyright notice may in some jurisdictions grant you _additional
> rights_ and help you claim damages, but not having it does not mean you
> forfeit your right of authorship.
> That said, since the notice may grant you additional protection, it's
> important to do it properly.  Specifically, beware that (c) (that is a c,
> in either case, inside parentheses) means nothing; either write "Copyright"
> out in full and/or (preferably and) use a real copyright symbol ©

I am not sure about using the copyright symbol alone, as it is UTF-8 only, 
while "Copyright" can be read by any ASCII-compatible encoding. (That is 
different from a book where the result is written and does not depend on an 
encoding to read.)

>
> So the format is simply this:
>
> © Copyright 2000-2002,2006 Your Name(s)
>
> > The internation conventions on copyrights, the local copyright
> > laws/systems... i.e. the same mess as with trademarks.
>
> It's not quite that bad :)  There are really only three cases to worry
> about: the ones that follow the Berne convention[1], the US[2] which is
> nearly but not quite precisely the same (and the US is in fact a party to
> the Berne Convention too, they just have some additional stuff mostly to
> deal with punitive damages - hence my warning about (c) vs © above), and
> then there are the  places that ignore copyrights entirely, and there's not
> much we can do about them.
>
> There are some excellent FAQ's around, google up Brad Templeton's FAQ if
> you're interested.
>
> [1] List is here, :
> http://www.wipo.int/treaties/en/ShowResults.jsp?lang=en&treaty_id=15
> The summarised version of the convention itself, if anyone is that
> interested, is here:
> http://www.wipo.int/treaties/en/ip/berne/summary_berne.html
> [2]http://www.copyright.gov/help/faq/faq-definitions.html#notice
>
> Regards,