Re: KDE not in Debian?
Marc van Leeuwen wrote:
> Andreas Pour <pour@mieterra.com> wrote
> > Marc van Leeuwen wrote:
> > > So we must "cause the App and X sources to be licensed as a whole at no charge
> > > to all third parties under the terms of this License". I've already commented
> > > on the word "cause"; in the ordinary sense of "cause and effect" this is not
> > > possible: as distributor, being author of neither X nor App, we have no say in
> > > third party's rights to the App and X sources.
> >
> > Right, but you forget about Section 7 of the GPL. It says there that if for
> > whatever reason -- including patent law -- you cannot comply with the GPL, you
> > cannot redistribute/copy. Since you cannot cause it to happen, you can't
> > redistribute: [...]
>
> The part about complying is obvious, this is the base of everything I've been
> saying. Yes, if you take "cause" to mean to take some action that achieves the
> stated effect and without which that effect would not have been achieved, then
> this is impossible, and the condition cannot be satisfied. But that would mean
> it can NEVER be satisfied, because the "you" in 2b does not own full copyright
> to the "work"; if he did there would be no need to abide by the GPL in the
> first place. So yes I do give a non-obvious meaning to this phrase. I did
> state that I was not happy with the formulation. Do you actually read what I
> write? I only bend the meaning here because the more obvious reading makes no
> sense: if the GPL is to contain conditions that can never be satisfied, it is
> useless, and that was not its intention.
Of course it can be satisfied. It's more difficult, obviously, when using a third
party's source, but if you are modifying a GPL'd program and the modifications are
yours this does not present a problem (you cause your own code to comply by releasing
it under an appropriate license and you cause the GPL code to comply by distributing
it, which by virtue of Section 6 of the GPL grants the appropriate rights).
Incidentally, you can also cause it to be true with third-party source, by taking
whatever action is required by the third party (which may be sublicensing, simply
distributing the third-party work or whatever) to cause the statement to be true.
However, I accept your interpretation, since I agree it is enough if the consequence
is true regardless of your efforts.
> So I am entitled to try a reading
> that is maybe a bit less obvious, but which makes sense out of the GPL.
>
> Besides, I wonder if my interpretation is so non-obvious. If by some chance I
> run into the obligation to "cause the sun to rise in the morning", wouldn't a
> judge find that I have fulfilled this obligation if in fact the sun did rise,
> without requiring me to demonstrate that I could have done anything to prevent
> it? Really this is all I've done, replaced a causal relationship by just a
> verication that the required effect does indeed take place.
>
> > > But those rights being fixed,
> > > we may check whether third parties do in fact receive the licence formulated
> > > in the GPL. For the App sources this is clear; for the X sources third parties
> > > in fact have licence to do anything the X licence permits, but since this
> > > includes all permissions given in the GPL, and not with stricter conditions,
> > > we may conclude that third parties receive full GPL licence with respect to
> > > the complete sources, and condition 2b is satisfied.
> >
> > Everything was going so well until you hit this point. In particular, the
> > statement "since [the X license] includes all permissions given in the GPL, and
> > not . . . stricter conditions, we may conclude that third parties receive full
> > GPL [license] with respect to the complete sources".
>
> As I have made abundantly clear
Guess it wasn't abundant enough -- where did you make it 'abundantly' clear?
> I interprete the verb "license" as "giving
> permission under conditions" and the noun "licence" as "conditional
> permission" (I've even tried to maintain the conventional British difference
> in orthography, but this subtlety was probably wasted on you);
Bloody right, as I'm not British, nor do I speak British English.
> just to be
> clear I've also added "(permission)" after "licence" at key points. The upshot
> is that I can give you a particular licence (permission) implicitly by in fact
> giving you a broader licence (permission). This is were the "no stricter"
> comes from, not from any particular words in any licence.
OK, so you did make it up. It's not in the GPL.
> So if I give you
> permission (license you) to distribute under the conditions of the X licence,
> I implicitly give you permission (licence you) to distrubute it under the
> stricter conditions of the GPL.
I thought we agreed we cannot change the license of the X code. Where is this coming
from?
Just to be clear, to distribute the X code under the "stricter" conditions of the GPL
means you have to forbid the recipient from doing a binary-only distribution (for
example) of the X code. So, as an example, if Debian distributes XFree code in their
distribution, then you would have to say that anyone who gets X code from Debian
cannot do what the X license says he can, namely, do a binary-only distribution.
Because if he could, then the X code is not licensed under the GPL, but under some
other license. Is that not so?
> > First of all, they do not in fact receive a full GPL license; if that were the
> > case, then they would not be able to make a binary-only distribution of the X
> > code which is distributed as part of the package, which they can under the X
> > license (which you agreed applies to the X code).
>
> How can having permission to do something (make binary-only distribution)
> prove I did not receive some other permission (full GPL)?
B/c GPL prevents you from doing it. So if you can do it, that means its not under the
GPL. Either that or the GPL doesn't mean anything.
> > As I see it, the phrase "licensed as a whole at no charge to all third parties
> > under the terms of this License" can have only two possible meanings. One is,
> > the whole must be licensed completely under the GPL. We both agree that does not
> > work with XFree code.
>
> OK you've managed to completely piss me off. I won't reply to or even read
> your messages any more, don't bother even to write them. I've spent some
> considerable time to give a very detailed description of how I read that
> phrase, and you elect to completely ignore that and just stick to your two
> readings of which the second is so absurd that I can't even start to
> understand it.
Well, sorry, I did not know you have a third one. What you say above, about implicit
permissions to distribute under stricter terms, does not make any sense to me, since
it is meaningless (if I have an explicit right to do a binary-distribution, I don't
see how your implicit permission for me not to be able to do it stops me from doing
it). And also I think it goes against what you said earlier, namely, that only the
author can change the permissions under which his work is distributed. So perhaps you
have been clear, but I don't understand you.
> > The other is, only the terms that apply explicitly to the entire whole apply to
> > the "added" code. One of those terms is in Section 2(b), which requires
> > distribution of the whole "at no charge to all third parties".
>
> Not "distribution at no charge", not even the GPL requires that. "Being
> licensed (obtaining permission) at no charge".
Right, I was a tad sloppy, that's what I meant.
> You are not allowed to require
> payment for giving permission (as you might if you own part of the copyright).
> Read. It's not difficult. Even by children can learn it. Just take your time.
Are these sorts of comments supposed to help the debate, or are you just trying to
insult me behind the shield of the Internet?
> > All of this would be pertinent if somehow the GPL distinguished between "more
> > restrictive" and "less restrictive", which it does not.
>
> Because it is implicit in the meaning of giving permission. Period.
That's not so obvious to me, at least if you mean by your statement that the greater
always includes the lesser (in fact the Supreme Court has often ruled that is not
so). And in particular it is not obvious with the X license, which requires you to
sublicense the X code under the XFree terms and not under other terms, such as the
GPL. I thought you had agreed to that (in your prior post where you wrote, "I do
agree with you that XFree (source) code cannot be licensed under the GPL" and "only
the author gets to state conditions for permitting distribution"), but I guess now you
don't anymore . . ..
To make this simplest, let's posit the situation which occurs in real life, namely,
the X source code is unmodified and is distributed with the GPL code. On the one
hand, in the prior message you seemed to indicate nobody could change the X code
license but the author, but now you seem to be saying anybody can make the license
stricter, and somehow this is supposed to be obvious to me.
Ciao,
Andreas
Reply to: