[conspire] d-check, minimal licensing

[Rick Moen]

'd-check' is Jesse Monroy's from-scratch rewrite of Ben Okopnik's
domain-check script in much, much cleaner Perl.  Now that Jesse's
finally settled on a licence (2-clause BSD), I'll be hosting it for
public download.

Thought some people might find this screed I just sent to be of interest.


----- Forwarded message from Rick Moen <rick at linuxmafia.com> -----

Date: Thu, 3 Mar 2016 21:39:07 -0800
From: Rick Moen <rick at linuxmafia.com>
To: Jesse Monroy <jesse650 at gmail.com>
Subject: d-check stuff (I forgot we had a discussion awaiting)

First, forwarding where we left 13 months ago.  Discussion resumes below
the cut.  (Yes, you settled on 2-clause BSD -- but I promised
explanation, so I'm supplying it, albeit a year+ late.)

----- Forwarded message from Rick Moen <rick at linuxmafia.com> -----

Quoting Jesse Monroy (jesse650 at gmail.com):

> I usually do BSD, but I see MIT around a lot. Can you point me to some
> literature on MIT?

I'll look around.  What do you want to know?  The MIT licence is
functionally identical to the 2-clause BSD licence.  I.e., it has the
exact same legal substance.  In case you hadn't made the connection,
it's the licence that the X Window System has always been under.


> I'd really like to put it as close to the public domain as possible.
> I'm not concerned if anyone make money with, or it's used as part of a
> product. My only concern is claiming copyright on it.

I'm sympathetic, but there's really no way around the fact of a
worldwide copyright regime built into the laws of every country and into
treaties.  Your work _is_ copyrighted whether you include a statement to
that effect or not.  Omitting a copyright statement merely creates
uncertainty for recipients.

----- End forwarded message -----

I respect people wanting to completely give work to the world, and make
copyright go 'poof'.  Best way I can think to cover this comprehensively
(in one go) is imaginary dialogue with 'JM'.  (JM isn't necessarily you;
it's a composite of people I've discussed this with.)


First, you asked about MIT License.  Executive summary from Wikipedia:

  The MIT License is a free software license originating at the
  Massachusetts Institute of Technology (MIT).  It is a permissive free
  software license, meaning that it puts only very limited restriction
  on reuse and has therefore an excellent license compatibility. The MIT
  license permits reuse within proprietary software provided all copies
  of the licensed software include a copy of the MIT License terms and
  the copyright notice. [...]

  The MIT License is similar to the 3-clause "modified" BSD license
  except that the BSD license contains a notice prohibiting the use of the
  name of the copyright holder in promotion.  [RM: 2-clause BSD License 
  omits this non-endorsement sentence, which is why I said MIT License is
  functionally equivalent to 2-clause BSD.]   [...]

  The MIT license specifically grants the right to "sublicense" in the
  text of the license itself -- a right not mentioned in the BSD license
  which simply grants the right to "redistribute and use." 

For more see former Open Software Initiative chief counsel Lawrence
Rosen's book _Open Source Licensing_, full text of which is online in
PDF format.  BSD licences and MIT License are covered in chapter 5, 
'Academic Licenses' (what I call 'permissive licences'):
http://rosenlaw.com/oslbook.htm

The imaginary conversation:


RM:  You should specify an explicit licence statement at the bottom of
this Perl script, though I certainly appreciate you giving me a copy.

JM:  Which licence?

RM:  Whichever one best suits you.  You're a BSD guy, so maybe you'd
like the 2-clause BSD license, though the MIT License is a bit better
drafted.

JM:  The part that really bothers me about this is claiming copyright at
all.  Can't I just put it in the public domain?  If not, can't I specify
licence terms and not include a copyright notice?

RM:  No, and no -- I'm sad to say.  Because it doesn't work.  Lots of 
people want to do this.  Their generosity does them great credit, but
the effort creates legal uncertainty for future users, and potentially
serious legal pitfalls.  

JM:  Why is that?  Shouldn't I be able to do with my property what I
want, including give it to the entire world?

RM:  Certainly you should be able to.  The reason you cannot is that 
huge, very rich copyright interests have altered copyright law in our
lifetime, to make copyright title arise automatically and vest in the
work's creator, in this case you.  

JM:  When was that?

RM:  In this country (USA), starting March 1, 1989.

JM:  What happened on March 1, 1989?

RM:  Congress's Berne Convention Implementation Act of 1988 took effect.
This was when the USA adopted the international Berne Convention, that
changed the way copyright works.

JM:  Are you saying it was possible to contribute works directly to the 
public domain in the USA before that?

RM:  Yes, though only a lawsuit and a judge's ruling could make it
definitive.  Basically, before 1989, if you published a work widely 
without any copyright notices and the matter came up in court, a judge
could rule that you had lost your copyright title through failure to 
assert it.  This point of law (the 'general publicaton' or 'divestitive
publication' doctrine) was important in BSD history as a turning point 
in the AT&T v. UC Regents lawsuit:  The US District Court judge's 1993 
preliminary ruling denied AT&T a preliminary injunction on grounds that 
AT&T had probably destroyed its own copyright under the doctrine of general
publication.  This was one of several reasons AT&T hastily settled 
immediately after that.

JM:  What change did the 1989 law make?

RM:  Basically, copyright now arises automatically for your creative
works, and persists until your copyright term expires, with or without
your inclusion of copyright statements.  The world's copyright barons
didn't want to lose copyrights accidentally any more, so they made it so
omitting notices can no longer nullify copyright -- whether you omit
them accidentally or on purpose.

People's vague memory of how things worked in pre-Berne times is part of
the reason many people think putting software up on the Internet without
mention of copyright or licences makes it public domain.  That was never
quite true (it required a judge's say-so), but is definitely never true
since 1989.

Also, before government rulings in 1974 for source code and 1983 for
binary code, the USA didn't classify software as copyright-eligible at
all, so that code (arguably) became public domain immediately.  Some
folks remember those days and think it's still like that.  It's not.

JM:  But surely I can still just say I don't want ownership.

RM:  You can.  But that is known to have zero effect in some legal
jurisdictions (e.g., the UK), and might have the intended effect in
some places, but it's totally up to a judge how to apply what you said
(if at all).  Unfortunately, in the post-Berne world, there's a very 
strong bias towards copyright title persisting and remaining or becoming
someone's property.  

JM:  But wait, you said copyright is now automatic with or without a
copyright statement.  So, copyright statements are now pointless, right?
Maybe I can achieve my intent by just omitting one.

RM:  Yes and no.  That would hide evidence of your copyright ownership,
but that ownership would still exist.  That would achieve one of your 
goals (issuing your script without annoying copyright cruft), but create
a huge long-term problem for recipients:  The wary ones would know
_somebody_ got copyright by writing it, and it's probably unexpired
(the only software with expired copyright terms being many decades old),
but wouldn't know who or since when.  The unwary ones might wrongly
assume nobody owns it.  This makes them unknowing copyright violators
who are courting a lawsuit (if an owner turns up), if they take actions
reserved to the copyright owner.

Basically, a work published without a copyright statement is a lurking
menace to recipients and reusers.  So, don't do that.  It's Not Nice.

JM:  What legal rights are reserved to a copyright owner?

RM:  For software works, the right to reproduce (copy) the work, the
right to make and distribute, sell, or give away the work or works
derivative of the work.  (For some other categories of work such as 
recorded music, there are additional reserved rights.)

JM:  Don't I convey some legal rights by just giving out a copy?

RM:  Yes, this is called an implied licence.  (You can grant licence in 
written form, orally, or through conduct.)  When you give me a copy of
your script, you are implying permission to run it, if necessary to
compile it, and so on -- whatever is required for me to implement your
implied intent in giving me a copy.  However, that does not include the 
right to redistribute it, or to make and distribute works based on it.
For example, my putting that script on my Web site for public download 
would violate your copyright rights, and you (or anyone who in the
future acquired your copyright title from you) could sue me for that
tort.

JM:  Even if I never registered the copyright?

RM:  Yes.  Unregistered copyrights are limited in what damages the owner
can collect, because the defence of innocent infringement is much more
credible and because of US statute law that limits damages.  But you can 
still get some damages and coerce an end to infringement.

JM:  Wait, I'm just doing a sanity check, here.  Are you sure it's not 
possible to put works directly into the public domain?  I've heard of
some efforts:  

1. Daniel J. Bernstein doing public domain dedications of qmail, djbdns
   and a bunch of his other works, at the end of 2007.
2. Creative Commons publishing a well-respected public domain dedication
   mechanism called CC0
3. WTFPL 2.0
4. Unlicense
5. A huge amount of old PD software such as Hunt the Wumpus, ADVENT,
   Super Star Trek, Maze War, Android, Rogue, Hamurabi, etc.
6. SQLite
7. CERN HTTPd following the CERN director's PD declaration of April 1993
8. David Goodger's 2002 Python tools for reStructuredText
9. The 334 projects currently on SourceForge with licence listed 
   as 'public domain'
   (https://sourceforge.net/directory/license:publicdomain/)
10. Works produced by the US Federal government

RM:  Taking those one at a time:


1.  DJB's mere 2007 statement that most of his projects are decreed public
domain raises exactly the long-term legal problems for recipients that I
mention above.  The best that can be said for certain is that Prof.
Bernstein gave everyone a powerful defence that would prevent him from
stopping or getting damages from users, redistributors, and creators of
derivative works.  

He has posted on his Web pages a number of reasons why he asserts that
his PD declaration has the intended legal effect.  I've dissected all of
them on my own 'public domain' page,
http://linuxmafia.com/faq/Licensing_and_Law/public-domain.html .  He
makes several key errors I discuss there:  One of the main ones is
confusing the enforceability of his copyright-owner rights (after an
overt act to abandon those rights) with whether he actually eradicated
the ownability of his software:  PD by definition has _no ownability_ --
by anyone, not just the author, but also heirs, creditors, and other
successors-in-interest.  

Dan leaves unsettled what happens in jurisdictions where PD declarations
turn out to be void by local law (such is known to be the case in the
UK).  He leaves unsettled what happens if his heirs, creditors, or
someone else comes into ownership of his property.  There's no reason to
think they're necessarily bound by his 2007 abandonment/disclaimer
statement.  And of course there are jurisdictions where PD-declaration
attempts have no effect, and may be others where a judge might apply
some third outcome.  The situation is largely indeterminate, and my main
point is this is an unkind problem to dump in posterity's lap.

(Dan or his heirs might also at some point invoke 'copyright recapture' 
and regain title as provided US law.  See:  
http://copylaw.com/new_articles/copyterm.html )

JM:  Wait, isn't Red Hat sponsoring a qmail fork?  Doesn't that mean
Red Hat, Inc.'s lawyers agree with DJB's PD assertion?

RM:  Respectively, sort of yes, and definitely no.  Yes, Red Hat
employee Prasad J. Pandit maintains N-DJBDNS aka ndjbdns (with all of
his additions under GNU GPL).  Last I heard, it's in the Fedora Extras
repo.  No, Red Hat Legal would have had no mandate whatsoever to decide
whether DJB's PD declaration is airtight.  Corporate counsel advise
management about whether the firm is assuming excessive legal risk.  The
most that can be said by us outsiders is that counsel probably advised
management that Red Hat, as a USA company, need not worry about ever
owing infringement damages to Dan.  This says nothing about whether
copyright title persists or what might happen if/when someone else
acquires that title, nor about what might happen outside the USA.


2.  Creative Commons CC0 is an excellent and effective legal instrument
-- but that's true because it doesn't rely solely on a PD dedication.
It does include one of those, an extremely well drafted one, _but_ CC0
also includes a fallback permissive licence that would take effect if
for some reason (such as in the UK), the PD declaration doesn't have its
intended effect.  

CC0 is thus an example of how to do this (PD dedication) right, if one
must do it.  (I should mention in passing that CC0 was created not for
software, but rather for scientific data.)  Ironically, the careful
drafting of CC0 also means it has all the legal features that PD
advocates typically deplore, including long legal language and a
copyright statement.

So, to sum, CC0 is excellent and highly reliable -- because it _isn't_ 
a PD declaration.


3.  WTFPL v. 2 is a 'crayon licence', i.e., incompetent.  Quoting what I've
written elsewhere:   

   Permissive licence so badly written that it grants rights only 
   to the _licence_ itself, and not to any ostensibly covered work.

Oh dear.


4.  Unlicense is another 'crayon licence'.  Quoting what I've written
elsewhere:

    Paragraph (and sentence) #1 professes to put the covered work into 
    the public domain.  Paragraph 2 professes to be a grant of rights 
    normally reserved by default to a copyright owner, which makes no 
    sense given that the preceding sentence professed to eradicate the 
    work's quality of being ownable.  _However_ (upon reflection), in 
    itself that would be harmless if redundant and pointless:  One can 
    interpret paragraph 2 as an elaboration of the consequences of the 
    first paragraph.

    Paragraph 3 is mostly further explanation of the concept of public
    domain, and therefore harmless if not useful.  Its middle sentence
    elaborates that the erstwhile author aims to bind heirs and
    successors, too (which is a logical inclusion, irrespective of
    whether it works).

    Paragraph 4, though, is the one that would be amusing if it weren't
    tragically broken:  It's the warranty disclaimer.  People accepting
    the covered work are obliged to accept the condition of no warranty,
    otherwise there is no licence.  Except, oh, wait:  Paragraph 1
    professed to put the work in the public domain, so the erstwhile
    owner has sawed off and evaporated in paragraph 1 all power to
    require the condition in paragraph 4.

WTFPL and Unlicense are leading examples of why clueless coders should 
not write your software licences, just as clueless lawyers should not 
write your code:  Either way, you're likely to get inadvertent comedy.


5.  Hunt the Wumpus and all that were mostly written before copyright
first applied to software at all in the USA (1975), and certainly before
Berne implementation (1989).


6.  SQLite, yeah, they claim it's PD by their own fiat.  And they
haven't sued anyone:  All of this tells you nothing at all about whether
copyright title got extinguished.  Additionally, the software (I think;
certainly their Web page) includes this statement, that to my eyes is a
perfectly valid open source licence:

  Anyone is free to copy, modify, publish, use, compile, sell, or
  distribute the original SQLite code, either in source code form or as a
  compiled binary, for any purpose, commercial or non-commercial, and by
  any means.

So, whether the PD dedication is legally effective will never arise in
court and get adjudicated, because there's a fallback permissive licence
-- exactly as with CC0.


7. CERN HTTPd, yes, they did.  Tim Berners-Lee referred to this:
http://1997.webhistory.org/www.lists/www-talk.1993q2/0259.html
It's unclear what the effect was under either Swiss (CERN's location) 
or US law, though.  Nobody ever litigated, and all of the covered code
was already becoming obsolete and has remained so.


8.  reStructuredText tools (http://docutils.sourceforge.net/), yes, 
Goodger intends dedication to PD.  Nobody knows whether this has the 
intended effect.  Also, there's a statement easily read as usable as a
fallback permissive licence, in the event of dispute:

  Dedicator recognizes that, once placed in the public domain, the Work
  may be freely reproduced, distributed, transmitted, used, modified,
  built upon, or otherwise exploited by anyone for any purpose,
  commercial or non-commercial, and in any way, including by methods
  that have not yet been invented or conceived.


9.  About the 334 projects currently on SourceForge with licence listed
as 'public domain', let me tell you about when I chatted with the
managers in charge of SourceForge when I worked at VA Software
Corporation (the company formerly known as VA Linux Systems but that
couldn't make up its name what it should be called):  I was building my
own catalogue of software for PalmOS for my Web site, and
licence-audited _all_ of the SourceForge packages with PalmOS relevance.

My findings included the fact that _every_ single package shown as
'public domain' or 'other' was erroneous in that claim.  In every case,
there was some undisclosed copyright encumbrance or licensing condition
that made the claimed status objectively untrue.

I showed management the details.  Their response:  Licence category is
self-reported information from the project maintainer.  VA Software
Corporation has no involvement, assumes no responsibility for its
accuracy, and would take no corrective action whatsoever.  End of
discussion.

The errors seemed to be inadvertent:  Basically, a lot of coders who
dislike dealing with legal matters just picked a category without
thinking.


10.  Yes, works produced directly by US Federal departments or agencies
go directly into the public domain by Federal law (17 U.S.C. 105).  
Oddly enough, this itself creates a problematic situation when such 
works go abroad, because the cited statute applies only _inside_ the
USA, and the Feds claim they are entitled to assert copyright elsewhere
(http://www.cendi.gov/publications/04-8copyright.html#317).  The
worst-case scenario is when a government work is published domestically
with no licence statement or copyright statement (because it's PD) and
then exported -- the 'lurking menace to others' problem I mentioned
before.

Government departments/agencies are being urged (see above link) to
include a statement like 'This is a work of the U.S. Government and is
not subject to copyright protection in the United States. Foreign
copyrights may apply.'  They aren't required to, but omitting this is
Not Nice.

Which underlies why you should do it, too, even though you are not
comfortable with 'claiming copyright'.


----- End forwarded message -----