Chris Nandor wrote:
>
>I have also included my reply to "Can we ignore licensing?" below.
>
>At 10:38 -0400 2000.09.11, Ben Tilly wrote:
> >1. Rephrase it so that it is clearly a copyright notice with
> >   an offered contract available for anyone who wishes to do
> >   things normally restricted by copyright.  Like the GPL.
>
>Why?
>
Because the things it wishes to do are best done through
contract law, not copyright law.  You can say anything you
want in a copyright, but if someone violates the permissions
you gave you are in a far poorer position to complain if you
do not have a valid contract to point to.

Specifically I have to wonder about whether any obligations
that are required are enforcable without a contract.  (And
yes, we do need a lawyer to answer that question.  IANAL and
my concern is drawn out of discussions about why the GPL is
structured that way.)
>
> >2. Freely Available is too vague.  Is it freely available if
> >   I release my changes in a form with a copyright notice
> >   saying (like Sun does) that you need to submit all of your
> >   changes to my changes back to me?  (Under the definition in
> >   the AL, yes.)
>
>If you can asnwer that question, then how can you say it is too vague?
>
Because vagueness has led to being overly permissive.

Take a look at how Microsoft "released" the changes that they
made to Kerebos.  Would releasing a changed version of Perl
under a BS license like that violate the understood spirit of
what the AL is supposed to protect?  To my understanding it
would.
>
> >   I think that 3a should be modified to
> >   something like "Released to a public forum under terms that
> >   are acceptable for reincorporation into the Standard Version.
> >   This means that it must fit all licenses that the Standard
> >   Version is available under.  Public domain will suffice but
> >   is not necessary."
>
>I think if Larry had meant that, he would have said that.  If that were in
>there, I personally would not use the "new" AL.
>
I think that Larry probably meant that.  What is the purpose of
making changes Freely Available?  Why obviously to allow Larry
to pick and choose what to put into the Standard Version of
Perl.

If you wish to make a public modified version and not let Larry
use your changes, the permission in 3c should be quite sufficient
IMO.
>
> >3. I think 3d should simply be "other arrangements."  Should a
> >   consultant build a custom version of Perl for someone else,
> >   distribution might not ever be an issue for either, yet they
> >   may wish permission to modify.
>
>That use is already covered under 3b.   The consultant is working for the
>organization in question at the time of the work for hire, and it therefore
>falls under "your organization."  Or you could flip it around, and say that
>the corporation itself is "you."  Either way it is fine and covered under
>3b.
>
If the consultant is an employee of a consulting firm I do
not know whether that argument would fly.  It might, IANAL.
But still there is no harm done by a more permissive wording
about something that is trivially true.
>
> >4) The permission for 4a) should only cover distributing the
> >   standard version.
>
>It explicitly only does.
>
No, but I think that is an oversight.

If you distribute two versions together, one of which is the
standard, you have indeed done "...at least ONE of the
following..." and therefore your distribution is legal.

It should be clear that you cannot distribute two together in
this dishonest fashion.
>
> >   As matters stand it is legal to have a
> >   distribution that distributes the standard version in some
> >   hidden place, but actually installs a piece of garbage.
>
>If it installs a piece of garbage, then it is not the Standard Version:

Of course not.  But if your distribution INCLUDES the Standard
Version then by my reading the distribution is legal.  And IMO
it should not be.

>       "Standard Version" refers to such a Package if it has not been
>       modified, or has been modified in accordance with the wishes
>       of the Copyright Holder as specified below.
>
>If you add more to it, then it is no longer the Standard Version.  If you
>change the installation, it is no longer the Standard Version.
>
"distribute a Standard Version".  You are doing that.

If it went to court I would not want to bet on your interpretation
winning over mine.  And as long as there is a question, the
wording deserves improvement IMO.
>
> >   Combined with an abuse of 3a), this could allow some nasty
> >   games to be played...
>
>What abuse of 3a?  If you distribute under section 3, you must make
>prominent what modifications you've made.  How can that be abused?
>
See Sun's Community Source License.  By reading the changes
you are bound by a nasty license giving them rights to
demand all sorts of crap from you.  Should distribution of
machine readible modifications with such a license attached to
decoding and reading them (which could, incidentally, make you
ineligible to contribute back to Perl) be acceptable?  Does it
interfere with the original author's artistic control?

I am not talking strictly hypothetical attempts at abuse.  I am
trying to adddress tactics already in use against free software.
>
> >5) I think that 4b is too dangerous.  Both Microsoft and Sun
> >   have played licencing games where you can see but not touch.
> >   Machine readable source that is combined with restrictive
> >   licensing does not meet the intent.  The purpose of this
> >   section is to allow ports.  (eg Macperl.)  I think this
> >   should have added that it must be distributed on the same
> >   licensing terms as the original.
>
>The Package must ALWAYS be distributed under the same licensing terms as
>the original.  Unless it is public domain or you are the copyright holder,
>you cannot change the licensing terms.  I am not sure what your point is,
>maybe I am missing it.
>
If I modify Perl I am now a copyright holder on my modifications
and I can distribute the alternate version with a license of my
choosing on my bits.

The point of the above is to insist that if you port Perl, your
port must be licensed on the same terms as Perl.  If you wish to
have different terms then don't call it Perl, don't cause any
confusion, and we will all be friends.
>
> >6) Item 8 belongs as a list item in item 4.
>
>No, because section 4 is about distributing in object or binary form, which
>is not necessarily what section 8 refers to.
>
Very good catch.  Section 4 should be any form then.  At the
moment I don't see permission to distribute an executable in
compresed form and that should not be an issue.  (In compressed
form you are neither object code nor executable.  Oops.)

IMO it *does* belong in section 4 and section 4 should relax
what forms of distribution are covered until it fits.

> >   The AL cannot
> >   redefine "distribute".
>
>It does not attempt to define it.  It attempts to say what shall be
>construed as distribution, not what will be defined as distribution.  And
>it can do that, and does do that.
>
Construed == what we define as distribution.

Under the semantics you are redefining distribution and I
would really want a lawyer saying that was an OK terminology
before I could feel comfortable with that phrasing.
>
> >6) Item 6 should include linked to this Package to extend it.
> >   Between when the AL was written and the present this has
> >   become a common thing to do (think modules), and should be
> >   explicitly allowed.  Note that this is not covered under
> >   item 7.
>
>Yes it is.  A pure-Perl module falls under "comparably compiled subroutines
>in other languages."
>
In general you are right.  I missed that.  But I should note
that there are modules that use Overload and would cause the
regression tests to fail.  But that is a boundary case which,
while it should be allowed, is not a big deal.
>
> >7) The GPL offers advice on how to apply it to your own
> >   software.  The AL should likewise.  In particular since
> >   the GPL is an offered contract, the disclaimer of warranty
> >   (section 10 of the AL, sections 11 and 12 of the GPL) does
> >   not clearly hold if incorrectly applied.
>
>I don't think that matters in the least.  So a user does not accept the
>license, but then sues because there is no warranty outside of the license?
>If that's what you're saying, it doesn't make sense.
>
All software comes with an implicit warranty that must be
explicitly disclaimed.  Which is why virtually all software
comes with an explicit disclaimer.

The GPL license likewise has a disclaimer, but since it is an
offered contract, in order to be valid there must be a quid
quo pro.  And therefore they make it clear that there is a
warranty to gain.

Hmmm...better answer than the above idea borrowed from the GPL.
Divide the licence into two parts.  The first part is the
explicit disclaimer of warranty.  The second part starts off,
"As part of this copyright notice there is an offered contract
for anyone who wishes to modify and redistribute..."

Now refusal of the contract part of the license does not void
the disclaimer of warranty. :-)
>
>At 10:52 -0400 2000.09.11, Ben Tilly wrote in "Can we ignore licensing?":
> >Suppose I buy a copy of Perl on a CD.  Under copyright law it
> >is fair use for me to resell that copy at *any* price I want.
> >That contradicts item 6 of the AL.
>
>You mean section 5?  No, it does not.  I don't see why you say that it 
>does.
>
Sorry, section 5.  Which says you cannot charge for the package.
And cannot charge an "unreasonable fee".
>
> >Incidentally items 3d and 4d effectively cannot be met if you
> >have many copyright holders (which Perl does).
>
>No, Larry is the Copyright Holder:
>
>       "Copyright Holder" is whoever is named in the copyright or
>       copyrights for the package.
>
>       "Package" refers to the collection of files distributed by the
>       Copyright Holder, and derivatives of that collection of files
>       created through textual modification.
>
>Copyright Holder refers to the copyright holder for the _package_ itself,
>not the individual pieces of it.
>
Copyright Holder is well-defined under law, and Perl may not
arbitrarily redefine it.

In addition Author says:

       Larry Wall <[EMAIL PROTECTED]>, with the help of oodles of
       other folks.

Under copyright law all of those oodles of other folks have
copyrights on the code and Larry cannot legally speak for
them.

Cheers,
Ben
_________________________________________________________________________
Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com.

Share information about yourself, create your own public profile at 
http://profiles.msn.com.

Reply via email to