CentOS - Apr 2015 - Real sh? Or other efficient shell for non-interactive scripts

On Mon, Apr 27, 2015 at 1:02 PM, Joerg Schilling
<Joerg.Schilling at fokus.fraunhofer.de> wrote:
> >
> The GPL makes claims that are in conflict with the law because these claims
are
> not amongst what the list in the law permits and that are thus void.
The GPL is all that  gives you permission to distribute.  If it is
void then you have no permission at all to distribute any covered
code.
> Both legal systems have the same results: They prevent the GPL from using
it's
> own interpretation os what a derivative work is and the rules from the laws
> apply instead.
So apply copyright law without a license.  You can't distribute.   I
agree that the FSF interpretation about distributing source with the
intention that the end user does the link with other components is
pretty far off the wall, but static binaries are clearly one 'work as
a whole' and dynamic linkage is kind of fuzzy.    US juries are
supposed to focus on intent and are pretty unpredictable - I wouldn't
want to take a chance on what they might decide.
> These rules make many combinations a "collective work" that is
> permitted. The cdrtools and ZFS on Linux match these rules - well, I assume
> that the ZFS integration code follows the rules that are needed for a clean
> collective work.
Can you point out a reference to case where this has been validated?
That is, a case where the only licence to distribute  a component of
something is the GPL and distribution is permitted by a court ruling
under terms where the GPL does not apply to the 'work as a whole'?
> Cdrtools follow these rules:
>
> -       No code from CDDL and GPL is mixed into a single file
How is 'a file' relevant to the composition of the translated binary
where the copyright clearly extends?   And why do you have any rules
if you think the GPL doesn't pose a problem with combining components?
  More to the point, why don't you eliminate any question about that
problem with a dual license on the code you control?
> -       Non-GPL code used in a colective work was implemented independently
>         from the GPLd parts and form a separate work that may be used
without
>         the GPLd code as well.
How 'you' arrange them isn't the point.  Or even any individual who
builds something that isn't intended for redistribution.   But for
other people to consider them generally usable as components in
redistributable projects  there's not much reason to deal with the
inability to   combine with other widely used components.   What's the
point - and what do you have against the way perl handles it?

-- 
     Les Mikesell
       lesmikesell at gmail.com

Les Mikesell <lesmikesell at gmail.com> wrote:
> On Mon, Apr 27, 2015 at 1:02 PM, Joerg Schilling
> <Joerg.Schilling at fokus.fraunhofer.de> wrote:
> > >
> > The GPL makes claims that are in conflict with the law because these
claims are
> > not amongst what the list in the law permits and that are thus void.
>
> The GPL is all that  gives you permission to distribute.  If it is
> void then you have no permission at all to distribute any covered
> code.
Fortunately judges know better than you....

If you read the reasoning from judgements, you would know that judges just look 
at the parts of the GPL that are not in conflict with the law. Judges know that 
making the GPL void as a whole would be a desaster.

> > Both legal systems have the same results: They prevent the GPL from
using it's
> > own interpretation os what a derivative work is and the rules from the
laws
> > apply instead.
>
> So apply copyright law without a license.  You can't distribute.   I
> agree that the FSF interpretation about distributing source with the
> intention that the end user does the link with other components is
> pretty far off the wall, but static binaries are clearly one 'work as
> a whole' and dynamic linkage is kind of fuzzy.    US juries are
> supposed to focus on intent and are pretty unpredictable - I wouldn't
> want to take a chance on what they might decide.
Given the fact that there is not a single trustworthy lawjer in the US that 
writes about the GPL and that follows your interpreation, I am relaxed.
> > These rules make many combinations a "collective work" that
is
> > permitted. The cdrtools and ZFS on Linux match these rules - well, I
assume
> > that the ZFS integration code follows the rules that are needed for a
clean
> > collective work.
>
> Can you point out a reference to case where this has been validated?
> That is, a case where the only licence to distribute  a component of
> something is the GPL and distribution is permitted by a court ruling
> under terms where the GPL does not apply to the 'work as a whole'?
There was no court case, but VERITAS published a modifed version of gtar where 
additional code was added by binary only libraries from VERITAS. The FSF did 
never try to discuss this is public even though everybody did know about the 
existence. As long as the FSF does not try to sue VERITAS, we are safe - 
regardless what intentional nonsense you can read on the FSF webpages.
> > Cdrtools follow these rules:
> >
> > -       No code from CDDL and GPL is mixed into a single file
>
> How is 'a file' relevant to the composition of the translated
binary
> where the copyright clearly extends?   And why do you have any rules
> if you think the GPL doesn't pose a problem with combining components?
>   More to the point, why don't you eliminate any question about that
> problem with a dual license on the code you control?
???

I completely follow the claims from both licenses, so there is no need to 
follow your wishes.
> > -       Non-GPL code used in a colective work was implemented
independently
> >         from the GPLd parts and form a separate work that may be used
without
> >         the GPLd code as well.
>
> How 'you' arrange them isn't the point.  Or even any individual
who
> builds something that isn't intended for redistribution.   But for
> other people to consider them generally usable as components in
> redistributable projects  there's not much reason to deal with the
> inability to   combine with other widely used components.   What's the
> point - and what do you have against the way perl handles it?
You are of course wrong and you ignore everything I explained you before.

If your idesyncratic GPL interpretation was true, your whole Linux distro would 
be illegal. When do you withdraw your Linux distro?

J?rg

-- 
 EMail:joerg at schily.net                    (home) J?rg Schilling D-13353
Berlin
       joerg.schilling at fokus.fraunhofer.de (work) Blog:
http://schily.blogspot.com/
 URL:  http://cdrecord.org/private/
http://sourceforge.net/projects/schilytools/files/'

Can we take the license wanking off the list please?  I don't think
either of the people arguing are actually lawyers, so it has no
relevance.
-- 
Chris Adams <linux at cmadams.net>

On Mon, Apr 27, 2015 at 2:13 PM, Joerg Schilling
<Joerg.Schilling at fokus.fraunhofer.de> wrote:
>
>> The GPL is all that  gives you permission to distribute.  If it is
>> void then you have no permission at all to distribute any covered
>> code.
>
> Fortunately judges know better than you....
>
> If you read the reasoning from judgements, you would know that judges just
look
> at the parts of the GPL that are not in conflict with the law. Judges know
that
> making the GPL void as a whole would be a desaster.
There is nothing in conflict with law about prohibiting distribution.
 And you cant' just unilaterally pick parts of the licence that
permits distribution that you like and ignore the rest.
>> So apply copyright law without a license.  You can't distribute.  
I
>> agree that the FSF interpretation about distributing source with the
>> intention that the end user does the link with other components is
>> pretty far off the wall, but static binaries are clearly one 'work
as
>> a whole' and dynamic linkage is kind of fuzzy.    US juries are
>> supposed to focus on intent and are pretty unpredictable - I
wouldn't
>> want to take a chance on what they might decide.
>
> Given the fact that there is not a single trustworthy lawjer in the US that
> writes about the GPL and that follows your interpreation, I am relaxed.
It's not 'my' interpretation.  Nor does my interpretation matter
much.
It is the owners of the GPL licensed code that would be allowed to
claim damages if the GPL terms are not followed.  And what they have
published is that all of the runtime linked components are included in
the 'work as a whole' specification.    I assume you are familiar with
RIPEM and the reason it could not be distributed until there was a
non-GNU implementation of  gmp.
https://groups.google.com/forum/#!topic/gnu.misc.discuss/4RcHL5Jg14o[1-25]

>> Can you point out a reference to case where this has been validated?
>> That is, a case where the only licence to distribute  a component of
>> something is the GPL and distribution is permitted by a court ruling
>> under terms where the GPL does not apply to the 'work as a
whole'?
>
> There was no court case, but VERITAS published a modifed version of gtar
where
> additional code was added by binary only libraries from VERITAS. The FSF
did
> never try to discuss this is public even though everybody did know about
the
> existence. As long as the FSF does not try to sue VERITAS, we are safe -
> regardless what intentional nonsense you can read on the FSF webpages.
Hardly.  One instance by one set of code owners has nothing to do with
what some other code owner might do under other circumstances. If you
could quote a decision that set a precedent it might be a factor.
>> > Cdrtools follow these rules:
>> >
>> > -       No code from CDDL and GPL is mixed into a single file
>>
>> How is 'a file' relevant to the composition of the translated
binary
>> where the copyright clearly extends?   And why do you have any rules
>> if you think the GPL doesn't pose a problem with combining
components?
>>   More to the point, why don't you eliminate any question about
that
>> problem with a dual license on the code you control?
>
> ???
>
> I completely follow the claims from both licenses, so there is no need to
> follow your wishes.
Unless, of course, you actually wanted the code to be used by others
or included as components of best-of-breed projects.
>> > -       Non-GPL code used in a colective work was implemented
independently
>> >         from the GPLd parts and form a separate work that may be
used without
>> >         the GPLd code as well.
>>
>> How 'you' arrange them isn't the point.  Or even any
individual who
>> builds something that isn't intended for redistribution.   But for
>> other people to consider them generally usable as components in
>> redistributable projects  there's not much reason to deal with the
>> inability to   combine with other widely used components.   What's
the
>> point - and what do you have against the way perl handles it?
>
> You are of course wrong and you ignore everything I explained you before.
And likewise you ignore the fact that you would not lose anything with
a dual license other than the reason for frequent arguments.  And my
only question is 'why not'?
> If your idesyncratic GPL interpretation was true, your whole Linux distro
would
> be illegal. When do you withdraw your Linux distro?
How so?   Which process links GPL and non-GPL-compatible licensed code
into a single work?   No one has suggested that it is a problem to
distribute separate differently-licensed works together on the same
medium or run them on the same box.

-- 
   Les Mikesell
     lesmikesell at gmail.com

On Mon, Apr 27, 2015 at 2:13 PM, Joerg Schilling
<Joerg.Schilling at fokus.fraunhofer.de> wrote:
> Les Mikesell <lesmikesell at gmail.com> wrote:
>
> There was no court case, but VERITAS published a modifed version of gtar
where
> additional code was added by binary only libraries from VERITAS. The FSF
did
> never try to discuss this is public even though everybody did know about
the
> existence. As long as the FSF does not try to sue VERITAS, we are safe -
> regardless what intentional nonsense you can read on the FSF webpages.
I just remembered a counterpoint to this.  Back in the Windows 3.0
days when windows had no tcp networking of its own, I put together a
DOS binary built from gnutar and the wattcp stack so you could back up
a windows or dos box to a unix system via rsh.    And when I tried to
give it away I was contacted and told that I couldn't distribute it
because even though wattcp was distributed in source, it had other
conflicts with the GPL.  As a side effect of getting it to build on a
DOS compiler, I prototyped the tar code and contributed that and some
bugfixes.  Someone else's version was accepted instead but at least my
name is still in a comment somewhere.  Probably the only thing still
being distributed...

-- 
   Les Mikesell
     lesmikesell at gmail.com