Re: [Gnu-arch-users] Re: arch and linux 2.7

[Stephen J. Turnbull]

>>>>> "John" == John Goerzen <address@hidden> writes:

    >> What makes you think that's not the license I was looking at?
    >> Those were exactly the words I was looking at.  I _still_ find
    >> no instances of the word "use" or its synonyms in that passage.
    >> Where are they?

    John> For me, and for many in the free software community, there
    John> is no clear distinction between "use" and "develop".

Nonsense.  There are 7000+ Debian packages.  In your life you will
continually qualify as a user of most of them.

    John> For instance, I do not consider myself an Arch hacker.  Yet
    John> I have contributed 64-bit patches, written three separate
    John> tla utilitity programs, etc.  My involvement with Subversion
    John> was similar.  From looking at this group, I'd say that this
    John> behavior is not uncommon from Arch users.

I'd have to say you participate in development.  Eg, I may very well
have a need for BK myself, and although I have never contributed a
patch to arch and have yet to publish my emacs mode archivist.el, I
will certainly buy a license.

    John> The vagueness of the clause above is enough to make this a
    John> violation.

Do you mean "broad"?  Yes, it's broad.  I think that broadness really
stupid from the point of view of PR, and IMO calls into question BM's
understanding of free software (more so than their support for it).

But not vague.  Some vagueness, there always is, but that vagueness is
a matter for the courts.  Despite your fears and any FUD that you may
have heard, Bitmover does not have the right to determine the meanings
of the terms "employee", "develop", "produce", "sell", "resell", or
"substantially similar capabilities".  If they claimed that right,
they would define them or reserve the right to do so, as they do with
the word "competing".  IANAL, but I'm will to bet my dollars against
your yen that _all_ of those terms have specific, well-known (to
lawyers in that specialty) meaning derived from case law and possibly
legislation.  In the case of the term "competing", Bitmover
_explicitly_ recognizes the primacy of the court (the word
"reasonable"), but asks the court to give it a fair chance against
products that harm its business while not being "substantially
similar" to Bitkeeper itself.

Of course the fact that this stuff is "well-known" only to specialist
lawyers gives Bitmover (and anybody who happens to retain a lawyer
specialized in contract law) an advantage.  That's life.

    John> Really, this means that

    >> Are you a lawyer, that anybody should give weight to your
    >> unsupported opinion about the "real meaning" of a legal
    >> document?

    John> No, but I'm not aware that anyone else in this thread is
    John> either.

I offer my opinion about the meaning.  Where I don't explicitly say
that it is my opinion, I have done research enough that I am willing
to risk eating my words, as I just had to on the issue of "IP ==
slavery."  I don't like the taste of bogosity-burgers.

But you simply don't understand the process of law at all AFAICT.
That's no sin, but to then go ahead and declare the "real meaning" to
be the most pessimistic interpretation you think you can get away with
is creating Fear, Uncertainty, and Doubt, no more and no less.

According to Tom this "cloud" of FUD is in itself harmful to the free
software community.  Think about that.

    John> if Larry decides you "compete", you are out.  So in the
    John> example of a Debian developer, Larry could decide that
    John> Debian competes, and take out all Debian developers if he
    John> likes.

    >> Wrong on four points.  First, you have implicitly equated "use
    >> other

    John> See above.

That's your problem; there is no ground for it in the law.  I get the
distinct impression that you think the rest of society should stop
what it's doing, and reorganize around the principles of free
software.  When put that way, it sounds absurd, doesn't it?  Certainly
not a good idea to hold your breath waiting for it.

In law, "develop" will have a specific meaning, distinct from "use".
If you wish to not make that distinction with respect to each package
you use, and still live in the general society, you will simply have
to consider yourself a developer, or make yourself a test case, and
contribute to the development of the law by defining one of the "edge
cases" for the user/developer distinction.  People (yea, even
programmers, All Hail djb!) do this all the time.

But whether you fall on the side of user or developer is your choice,
and you only have to worry about it with respect to SCMs and
"competing products".  If you don't like that, don't use BK under the
"free use" license.  That the BKL is unfree is a valid complaint.  But
don't say that the BK "free use" terms gives Larry a hunting license
for Debian developers.  It doesn't.

And, of course, this has nothing to do with being a member of a
particular distro's developer crew: _all_ free software users are
subject to this difficulty.

    >> Second, with respect to the more plausible contention that
    >> "Debian developers" en masse are disqualified because Debian
    >> distributes CVS and svn, that license does not allow Larry to
    >> decide that _you_ compete.  It allows Larry to decide that a
    >> _software product_ competes.  He must then establish that you
    >> or your employer are "developing, producing, selling, or
    >> reselling" _that software product_ or he may not invoke
    >> paragraph 3(d).
    >> 
    >> It might actually be non-trivial to prove that Debian as a
    >> distro develops, produces, sells, or resells other SCMs in the
    >> legal meaning of those words; I don't know.  It's surprising
    >> how often the law

    John> Then we have to err on the side of caution.  Even if that is
    John> the case, a lengthy court battle with Larry would probably
    John> be as expensive, if not more so, as actually buying the
    John> comercial version, so he wins either way.

*chuckle*  You _are_ naive.  Nobody wins a court case except the
lawyers, any more than you can "win" a nuclear war.  Unless Larry is
vindictive beyond your imagination, he'll settle for revoking your
license.  Unless the case has deterrent effect hundreds of times
bigger than any likely damages he'll get.  Ben Collins being a case in
point -- BM possibly could have got statutory damages for that, looks
like a case could be made for intentional infringement.  But they
settled for license revocation, didn't they?  Doesn't sound terribly
blood-thirsty to me.

But _you_ don't have to err on the side of caution, unless you're an
employee of Debian.  If you are not an employee of Debian, then it's
much more clear-cut.  (Exception: you might be an independent
reseller.)  You personally either develop a competing product, or not.
Most Debian developers don't work on SCMs ever at all, so they have no
problem.

    >> Third, you are ignoring the word "reasonable".  That word is
    >> defined by the courts and any applicable legislation, not by
    >> Larry or Bitmover's lawyers.  Ie, Debian undoubtedly produces
    >> configuration

    John> It is?  Exactly how is it defined?  To me, that sounds
    John> extremely vague -- as long as they can come up with a
    John> justification, they're OK.

First of all, the point is mostly moot.  Nobody here would expect
anything but a reprimand for wasting the court's time for claiming
that arch or svn is not a competing product.  "Competing product" is
the only term that Bitmover claims the right to define, so there is no
conflict, and no need to resort to "reasonable opinion".

The point of the word "reasonable" is that you (and the court) should
be able to predict what Bitmover will judge to be a competing product,
based on standards held by "reasonable" well-informed people.  If by
"vague" you mean "broad", yes, Bitmover will make a broad claim.  But
there are "well-known" (once again, to specialists) limits to that
claim, so you can figure it out in advance, and know whether you're
about to do something that would violate the license.  Of course,
lacking legal expertise you should err on the side of expecting
broader claims, or get a lawyer's opinion.

There _are_ standards in law for defining competition.  One is the
"substantially similar capabilities" definition.  Arch obviously
satisfies that.  Others are negative.  For example, "autoconf" is not
a competitor to BK, although it manages software configurations.  Most
important uses of BK (a plausible sample would be everything in
bkbits) are going to use autoconf _with_ BK.  I'm pretty sure that a
court would look at that fact and throw out the "competitor" claim.
(In anti-trust law that is definitely true.)

In between, the court will put a lot of weight on what Bitmover
claims; that's part of contract law, determining the intent of the
parties.  By accepting the license, you submit to Bitmover's
reasonable opinion.  But you can contest whether Bitmover's claim is
"reasonable", for example by calling expert witnesses from the field
who would say "no expert would consider that a competing product."
I'm pretty sure (as a layman with some related expertise) Debian would
win if BM claimed dpkg was a competing product.

Once again, there is distinct advantage to the side equipped with
lawyers ex ante, and if it comes to court the deeper pockets have a
huge advantage.  But there's just no way to avoid that.

    >> Fourth, to the best of my knowledge, Debian has _no_ employees.
    >> If so, only Debian developers directly involved in the SCM
    >> packages are at any risk of being denied use under the
    >> soi-disant "free use" license.

    >> Combining those four points, I think that Debian users are at
    >> zero risk of not qualifying for the "free use" license, and the
    >> risk to Debian developers in general infinitesimal.

    John> We have one Debian developer, Ben Collins, that has already
    John> been threatened by Larry and had his Bitkeeper free use
    John> license revoked.

Put down your outrage and look at the facts.

Ben Collins is not a "generic" Debian developer for the purpose of
this license.  He's a core developer in svn and was explicitly working
on bk2svn capabilities, using bk (so he probably violated 3(g)(iii),
too).  It took me less than five minutes (including a detour to
www.subversion.org) to confirm those things.

What, BM announced the rules, and you are surprised that they actually
intend to play by them when confronted with an egregious violation?

    John> Moreover, there is another danger: if I work for a large
    John> company, they may be developing products that compete with
    John> Bitkeeper in some other department that I don't even know
    John> about.  I could then be liable for a $100,000+ license
    John> violation fee, or have my free use license revoked.

Yup.  So what?  It's horrible, but irrelevant to your claim that
Debian developers in general are at risk.  Get a more transparent
employer if that worries you; there may be such clauses in the BIOS of
your workstation.  N.B. Claiming you are actually liable for the "up
to $150,000" statutory damages is probably FUD (much of it Bitmover's
FUD).  Here's the law:

------------------------------------------------------------------------
[Somewhat abridged from http://www.copyright.gov/title17/92chap5.html#504]

(c) Statutory Damages.

(1) Except as provided by clause (2) of this subsection, the copyright
    owner may elect, at any time before final judgment is rendered, to
    recover, instead of actual damages and profits, an award of
    statutory damages for all infringements involved in the action,
    with respect to any one work, for which any one infringer is
    liable individually, or for which any two or more infringers are
    liable jointly and severally, in a sum of not less than $750 or
    more than $30,000 as the court considers just. For the purposes of
    this subsection, all the parts of a compilation or derivative work
    constitute one work.

(2) In a case where the copyright owner sustains the burden of
    proving, and the court finds, that infringement was committed
    willfully, the court in its discretion may increase the award of
    statutory damages to a sum of not more than $150,000. In a case
    where the infringer sustains the burden of proving, and the court
    finds, that such infringer was not aware and had no reason to
    believe that his or her acts constituted an infringement of
    copyright, the court in its discretion may reduce the award of
    statutory damages to a sum of not less than $200.
------------------------------------------------------------------------

Section 504(c)(1) says that _total damages_ for _all_ infringements on
a single work are in the range $750--$30,000, and the court decides.
I can't imagine the court would consider it just to inflict the larger
end of the range on you in the case in question.  Granted even $750 is
a lot of money to you, Bitmover wouldn't be willing to spend an extra
hour of billable time to get that tiny amount of money.  Actual
damages wouldn't be very big, either, the cost of a single seat
license, and only if the court was convinced that without access to
the free license you probably would have purchased a license (the
court would not automatically assume you would buy a license in the
inadvertant case).

The first sentence of Section(c)(2) rules out the inadvertant
violation you describe, and plaintiff must "sustain the burden" of
proving it.  So you can even avoid the $150,000 penalty simply by
saying "yeah, I did it on purpose".  Or you can contest it and
probably (since we're assuming the truth is that it was inadvertant)
get the statutory damages reduced to $200.

Bitkeeper would be nuts to do anything but get you to agree to
revocation of the license.  Tell your boss you want a raise to cover a
commercial license, or you'll go work somewhere where you can get the
free license.  *shrug*

========================================================================

Look, I don't support this license.  Clauses 3(d) is hideously and
arbitrarily (from the point of view of the user/contributor)
discriminatory in the free software context.  There are serious
problems with it, over and above it simply being non-free from the
get-go.  But your claims about the impact on distribution developers
and its vagueness (unless that's a typo for "broadness") are simply
unwarranted FUD.

Just point out the real problems, instead of spreading FUD:

0.  It's not free.

    * If Bitmover were genuinely supportive of free software, it would
    make an effort to provide a free version.

1.  (As pointed out by Paul).  The lack of a published version of the
    license on the website forces those who wish to understand its
    terms and how it might apply to their situation to download and
    run the software.  To a non-lawyer, this feels entrapping.  (It
    might actually be entrapping, I don't have any knowledge about it.)

2.  The license, section 3(d), makes a distinction between users and
    developers which is (from the proprietary point of view) legally
    fairly clear and useful.  However, in the context of free software
    development, it is more ambiguous, and probably very aggressive
    about assigning developer status.  Therefore it severely inhibits
    the easy transition between user and developer that is the
    hallmark and raison d'etre of free software.

3.  Section 3(d) also makes individual developers liable to
    inadvertant violation due to employer behavior.  In the free
    software culture, this effectively puts all employees of large
    companies at risk, again chilling the natural participation by
    users in the evolution of the software they use.

    * If Bitmover were genuinely supportive of free software, it would
    recognize facts 2 and 3 and use different methods to protect its
    copyrighted assets.

4.  Section 3(f) requires the user to stay current or lose the
    license, but does not provide means for the user to determine
    whether an update is required, a reasonable grace period for the
    update, or means to access user data in the event that the update
    should have a defect preventing access.

    * This is just plain buggy.  Bitmover should fix it, by undertaking
    to notify registered users by email and on the website, and by
    providing a grace period long enough to update the software and
    verify its continued correct operation.

There are likely others, but these are the ones that have come up in
this thread that I've noticed.

-- 
Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
               Ask not how you can "do" free software business;
              ask what your business can "do for" free software.