[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: discussion with the FSF: GPLv3, GFDL, Nexenta

On Sun, Jun 03, 2007 at 12:28:04AM -0700, Don Armstrong wrote:
> If the author of Star decides that the Debian maintainer has
> incorrectly removed a "copyright notice",[1] he could terminate the
> license under 6.1, 

He could claim the license is terminated under 6.1, but presumably the
Debian maintainer would dispute such a claim.

> and bring action in Berlin for copyright
> infringement; the maintainer and any other parties to the action
> (people to whom the work was distributed after notification of breech)
> would then have to defend themselves in Berlin instead of notifying
> the court that the venue was improper (or whatever the German
> equivalent is.)

The court in Berlin would have to not throw the case out on their
own accord (in spite of the difficulty in having your side of the case
presented, and in spite of the jurisdictional issues, the questionability
of the claim in the first place, and the difficulty in showing harm),
rule against you in absentia (agreeing with the arguments presented),
and could then only take action in so far you already operate in its
sphere of influence, or in so far as it can convince your government to
extradite you or enforce its rulings for them.

Should someone be willing to do that, and a court is willing to go
through all those steps with a choice of venue clause, what makes you
think they'd not do so in the absence of one?

> On Sun, 03 Jun 2007, Anthony Towns wrote:
> > > Since this is giving up a right normally enjoyed in exchange for
> > > the ability to use or modify a work, it appears be a fee, and as
> > > such fails DFSG 1.
> > You're not giving up any rights, you're gaining the right to modify
> > and distribute the software under certain conditions, just as you
> > are under the GPL.
> We don't give up rights under the GPL that we otherwise enjoy though;

Some people do. Microsoft considers the right to enforce its properly
obtained patents worth going to the trouble of distributing coupons
instead of SuSE itself, eg.

> we only gain ones in specific circumstances. In the case of the CDDL,
> we lose rights even in the case where we're only using the work.

What makes you think the latter is true? I don't endorse the claim that
copyright licenses can take away usage rights if you're not making use of
the ability to modify or distribute that they offer you. In some cases
it may be enough to provide a simple notice like that to bind a user,
but that's dependent on your jurisdiction as a user more than a "choice
of venue clause", and I can't see any reason to think it applies to the
CDDL even so.

> > You're required to give up something you might value and otherwise
> > demand compensation for, certainly, but there needs to be something
> > more than that to violate the DFSG.
> "giving up something that you might value [or] otherwise demand
> compensation for" applies equally well to cash money as it does to any
> other intangible which has value. A requirement to send an email to
> the licensor if you possibly can isn't "cash money" either, but it
> sure seems to be a fee to me.

It's not a fee in the normal sense of the word, but it is a restriction
in the sense that if you're not able to do it (and you may well not be
able to), you're not able to make use of the priveleges you're offered
in return. That's where the analogy to a fee comes in -- it stops some
people from being able to participate.

For a choice of venue clause though, it only stops some people from
being willing to participate; just as potentially giving up patent rights
stops Microsoft from being willing to distribute Linux.

It's *possible* that it's still obnoxious enough that it's too much to
ask, but so far I can't see any significant cost to choice of venue that
makes it any worse than all the other weird and wacky things people put
in free software licenses.

> > The DFSG are a set of *guidelines*, if you can't explain violations
> > in simple, understandable terms, they're not violations.
> This is my understanding as well; I'm only explaining the application
> to DFSG 1 to attempt to appease strict constructionists.

The OSI lists are that way: ---->


> > > I'm personally using "feel" as shorthand for "my understanding of
> > > the legal situtation regarding this clause and its relation to the
> > > DFSG"
> > That's great, but *your understanding* isn't any more important than
> > anyone else's.
> I'm not claiming that it is; my point is that my understanding is not
> *less* important than anyone else's. I've done what everyone should do
> to come to an understanding.

I'm glad to see you write that; though I was referring more to Francesco's
post and similar than yours. 

But even so, when you say things like "I'm personally more concerned
about licensing than the average developer" and "I [...] expect people
who disagree with my analysis to actually engage the analysis with
counter arguments, come to a complete understanding of the problem,
and then make a determination" you are saying your understanding is more
important than other people's.

Holding people who agree with you to that standard might be a way
to start?

> > There's something fundamentally wrong with the way discussions work
> > on debian-legal that people think that simply posting their
> > understanding is a valuable contribution.
> What else can we do? We take input, we examine it, we respond with our
> understanding of how the input meshes together. I don't believe we're
> capable of presenting absolute truth.

Who is "we" in the above? For someone who's not a regular on -legal,
it doesn't sound like it includes me.

In any event, the important thing (afaics) isn't to have a forum where
regulars can post their understanding of issues, it's to help the people
you're communicating with have a better appreciation for the complexities
involved in their issue and how they might choose to approach them. That
can mean pointing out possible drawbacks in existing licenses, explaining
tradeoffs between licenses, or suggesting alternative ways of drafting
licenses that avoid having to make some tradeoffs, but it doesn't mean
making the tradeoffs for other people.

> Or we can try to understand the basis for our positions, and either
> come to a place where we agree, or have completely plumbed the
> argument so that we agree to disagree. This is my goal. I don't
> believe we've come to this position on the CDDL yet.

At present, the important drawbacks to the CDDL ttbomk are:

	* it's YA license
	* it's GPL incompatible

the benefits are:

	* it's a license that allows free use, distribution and
	  modification that Sun's comfortable using for some interesting
	  software, that's not otherwise available under a free license

and a vaguely interesting note is:

	* actually suing based on the license might be complicated by a
	  choice of venue

That you can argue the latter is analogous to a "fee" isn't really
very interesting. That some people are concerned about it is more so,
though so far the only concrete concern I've seen is MJ's comment --
"A possible arbitrary lawyer-fee-bomb, depending on the venue specified
and its sanity." and that's not really very concrete either.

> If we end up agreeing to disagree, then we should punt, and use the GR
> process to decide whether the work goes in main or not, and have each
> side write up a explanation of the problems surrounding the license,
> and publicise it with those that agree signing on to it.

No, punting to a GR is not a good solution -- it's slow to come to a
resolution, it annoys developers who have to inform themselves about
something they'd rather not worry about, and it ends up with -legal
folks complaining that the resolution doesn't make sense.

It's also not called for here -- we have CDDL software in the archive
and have notes from an ftpmaster and the DPL that it's acceptable. If
you want to change the current situation, you need to convince people
it needs to be changed.

(You'll note that in the GFDL case, changing from the then current
situation of not worrying about GFDL stuff had the support of all of
ftpmaster, the DPL, and the release managers amongst others)

> > implying that other people aren't sufficiently "concerned about
> > licensing matters", aren't "actually engag[ing] the analysis with
> > counter arguments", don't have a "complete understanding of the
> > problem" in order to stop them "mak[ing] a determination" sounds
> > like a pretty good match for the last case.
> It's very much the opposite, actually. I'd like nothing more than to
> have people who disagree with my understanding attack the rationale
> behind my understanding so at the end of the day, no matter how wrong
> my understanding starts, I end up getting things right. I like to
> believe that many -legal contributors have this mindset as well.

Well, what you want and what your actions are likely to result in are
often two different things :)

> > Ultimately Debian's policy isn't going to be decided by whoever
> > understands legal issues the best, it's going to be decided by the
> > developers who contribute to Debian, whether they fully understand
> > things or not. Trying to limit the discussion to experts is all very
> > well, but it'll just leave non-experts ignoring the discussions when
> > they end up making the ultimate decision.
> So what should we do? I can't force people to become experts, and I
> very much want to avoid bludgeoning people with licensing issues who
> aren't interested in hearing about them.

So talk about things on the level that they actually care about, rather
than the level you're most comfortable with.

For me, personally, that's practical problems that get in the way of
developing and promoting awesome free software -- and at the moment,
debian-legal rates far worse on that score than anything I've seen about
the CDDL does...

> I don't believe I've ever intentionally advocated my understanding or
> opinion as an official consensus, and I don't plan on starting now. In
> fact, when I talk publicly about such things, I'm very careful to
> indicate that my opinion is nothing more than my own. Insinuating that
> I would intentionally mislead users is rather insulting, frankly.

From your original post in this thread:

] >     3. Nexenta: Despite their incompatibility, Debian accepts both the
] >  CDDL and GPLv2 as valid free software licences and would welcome any
] >  solution to the distribution of a Debian system based on OpenSolaris.
] This is not the case, unfortunatly, and it really would be wise in the
] future to consult with people who are familiar with the arguments
] surrounding such licenses before expressing Debian's opinion to the
] FSF.
] The CDDL's clause 9 is very much not appropriate for works in main,
] and to the best of my knowledge, works licensed solely under the CDDL
] have never been accepted in main.[1]

Can you see how someone might mistake those comments for something more
than just your personal opinion?

In particular compare saying "This is not the case" with, eg,
"I disagree".

(I'm not intending to imply that you're intentionally misleading anyone,
or that anyone else on -legal is. Personally, I'd say it's more an
artifact of the force with which we're in the habit of presenting our
views on Debian lists; but whatever the cause or intent, the problem's
in the effects.)

> Getting back to my point: users should be aware that CDDL licenses
> with this clause may be result in giving away rights that they may
> otherwise not want to give away, not that Debian's offical opinion is
> that CDDL licenses are bad. This is the area that education is useful,
> just as explanations to users why GFDL works even without invariant
> sections still have problems that they may want to be aware of, even
> though Debian's official opionion is that they aren't serious enough
> to exclude them from main.

Right -- but the latter is an example of why we might well want to keep
CDDLed software in main, even though we might also want to help users
or redistributors understand potential problems they might face.

And again, I don't think I've seen any real statement of what those
problems actually *are*.

> What would make it more welcoming? A large part of the problem is the
> need to continuously point out counter arguments, [...]

What makes it unwelcoming is the appearance of a consensus that doesn't
brook argument, even when that consensus differs significantly from that
of other sections of the free software (or open source) community.

> > In particular, whoever ends up responsible for Debian's official
> > policy will need to spend their time educating users on what the
> > official policy actually *is*, not their opinion on what the
> > official policy should have been.
> We have very few "offical policies" though; 

Uh, I don't know who you mean by "we" there -- but Debian has many
official policies, even just counting the ones of the form "license
xyzzy is (not) DFSG-free".

> there are some things that
> are widely agreed to be free, some things that are widely agreed to be
> non-free, and some things which are much more contentious. Someone who
> is capable of talking about the official policy is going to
> necessarily have their own opinions and will express it.

I don't think that actually makes sense -- Debian's policy on this is
expressed by what will actually get accepted into the archive (or what
will get removed for license reasons). That's decided by consensus
(where it's obvious), by ftpmaster (where it's not entirely clear or
somewhat controversial) or by general resolution (where it's significantly

But at that point Debian's policy on whether a given license meets the
DFSG is one of objective fact -- Debian does accept the GFDL as a free
license (ignoring invariant sections), and Debian does accept the CDDL
as a free license (at least when the choice of venue is Berlin).

You can entirely reasonably have opinions on whether there are better
ways of getting what you want, better licenses, whether the license meets
your personal standards of freedom, whether Debian's standards should be
different, whether this policy seems contradictory to some other policy,
or whatever else.

But the question of whether something is DFSG-free is fundamentally one
that's resolvable, and it needs to be seen that way.

> > To take a particular example: if you want to retain the privelege to
> > call the GFDL vote result "wrong", you're excluding yourself from
> > being in a position to define Debian's interpretation of "free
> > software".
> Not at all; it's quite possible to publicly state that a ruling was
> incorrect and still adjudicate under the ruling. 

If you want to be trusted to make similar rulings on behalf of the people
who intentionally made the decision you're calling "incorrect", then no,
it's not possible.

> > Personally, I think comparing the reactions to getting overruled by
> > GR of ftpmaster members to various subscribers of -legal is probably
> > instructive.
> Like whom? While many may have expressed dissatisfaction with the
> outcome,[2] to my knowledge no one on -legal has expressed as the
> "offical policy" on the GFDL contrary to the GR outcome without being
> corrected.

No, the reaction of saying things like:

] From: Anthony DeRobertis <anthony@derobert.net>
] Subject: Re: Results for Debian's Position on the GFDL
] Date: Sun, 12 Mar 2006 17:15:40 -0500
] [...] 
] I agree wholeheartedly, but I'm not exactly sure how else to proceed. I
] don't think /anyone/ who is part of the consensus here is too happy with
] the outcome of this GR.
] Alas, now that pi != 4*atan(1), how shall we proceed? Interpreting
] licenses and the DFSG is nowhere near as clear as mathematics and,
] unfortunately, just ignoring the GR would, I think, make us look like
] sore losers.

because clearly everyone who voted for the winning option is the sort
of person who would think pi can be redefined willy-nilly, or that the
only reason to respect the GR is to avoid looking like sore losers...

(Although while Anthony DeRobertis was a -legal regular at the time,
he doesn't seem to be a DD, maintainer or n-m applicant afaics. Does
that mean he's not a part of the consensus on -legal?)


Attachment: signature.asc
Description: Digital signature

Reply to: