For Approval: Microsoft Permissive License

Chris Travers chris.travers at gmail.com
Mon Sep 10 20:34:02 UTC 2007


I post this in the hopes that it helps people other than Rick since I know
he doesn't read my posts.

On 9/10/07, Rick Moen <rick at linuxmafia.com> wrote:
>
> Quoting Ben Tilly (btilly at gmail.com):
>
> > It may be clear to you, but it is not clear to several others,
> > including me.  Multiple people have pointed out the obvious case of
> > Tivo, if they had a Linux kernel patch, would they have cause to
> > object if the kernel went from GPL v2 to GPL v3?  They contributed
> > expecting to be able to benefit from future improvements made by
> > others, and now they can't benefit.
>
> 1.  Have you confirmed that there is in fact code in the Linux kernel
> that bears a TiVo, Inc. copyright notice?  Looking through
> http://www.kernel.org/pub/linux/kernel/CREDITS , I see nothing even
> remotely like that.  (Yes, I know you said "if".  I am in general less
> interested in hypotehtical situations than in real ones, particularly
> when other people are invoenting [sic] work for me regarding those
> situations.)


In my project, we do occasionally forget to cite patches.  At any rate,
without doing an exhaustive search for patches, it is difficult to know.



(If prior contributors _meant_ to speak of TiVo as a kernel contributor,
> I apologise for having missed the intended context, which was very
> unclear from the wording of their posts.)



THe context was missed due to Rick's procmail filters. :-)

2.  It is not clear to me that Torvalds as project leader has a legal
> duty to preserve the ability of contributors (or other users) to
> continue to make a particular economic use of the codebase (especially
> as they can fork at the time of any such licence change).  If you
> believe doing so is a tort on the part of the project leader towards
> (hypothetical) contributors, kindly tell us what tort that would be.


Does it cause harm to copyright owners?  You seem to suggest that such is
the bar.  Causing a contributor to lose an ability to make a certain use of
the codebase would seem to cause economic harm.


> Similarly for many embedded devices, the length of the GPL v2 is
> > already an issue.  The GPL v3 is much longer.
>
> Again, if you believe a tort would be committed, please feel welcome to
> cite what specific tort that would be.



If it drives up costs and hence drives down revenue, that would be a
quantifiable damage, would it not?


> As a practical matter, very few open source projects have properly
> > registered copyrights to deal with.
>
> Correct.  I note in passing that this concern may be US-centric.  I do
> not know how many other jurisdictions give greater enforcement power to
> registered copyrights than to unregistered ones.
>
> > And so it may be hard to quantify and prove actual damages.
>
> Actually, it would be difficult to quantify and prove actual damages
> _irrespective_ of whether the copyrights in question had been
> registered.  Lack of registration, however (in US jurisdictions), makes
> such a showing pointless (until after registration), as collection of
> such damages is barred by statute.


Wrong for reasons listed below.


> But there are real drawbacks for some contributers to moving to the
> > GPL v3.
>
> Immaterial to the legal question, as stated.  Again, if you claim that
> tort theory applies, kindly do detail what torts, how, when, etc.



IANAL, but I think your characterization is dangerous.  As I understant it,
all that one is required to force a trial is to show:
1)  That one is able to articulate a claim under tort theory.
2)  Able to articulate facts which, if true, would make the claim actionable
(i.e. that one has been harmed *in some way* by the claim).

In short, the bar to forcing a trial is far lower than what is required to
win it.  After all, the trial is there to determine what the facts of the
case actually are.

> Whether or not they would win the legal case, the objections are
> > real....
>
> "Objections" are not the same as civil wrongs.  Please do not confuse
> the two.



Ok but as I pointed out before, objections can cause harm to a project which
can cuase financial damages to those providing support.

> Suppose that I think a licensing change benefits 95% of my
> > contributers, but is bad for 5%.  Is it OK to make that licensing
> > change?
>
> I would imagine that a project leader would wish to look out, mostly,
> for the project as a whole.  However, if he messes up regarding 5% of
> contributors, he/she has to bear the risk of them being angry, and had
> better have a compelling reason for his/her actions -- and, separately,
> if the complainers feel wronged as a matter of civil law and are
> sufficiently highly motivated, the project leader could end up
> attempting to convince a court otherwise.



ANd that is a massive issue, when you think about the amout of lost time and
effort that would come with that.  I am not saying "never"  but I am saying
that it should take a clear and existential threat to a project to make this
happen ethically and for the good of the project.

This state of affairs, of course, is indistiguishable from pretty much
> the entire rest of life in a real society.



Agreed, why I say that stability is an important good, and that changes are
painful :-)

> I submit that it is not, because I have no right to coerce
> > that 5% to give of their freely given labor what they didn't choose to
> > give.
>
> I submit that trying to keep everyone absolutely happy with one's
> actions, all of the time, is always and everywhere a fool's game, and
> has been universally known to be so since _at least_ the days of Aesop.



True.  But when in doubt, you keep things the way they are until you have a
clear necessity for change.

> YMMV
>
> Damned straight.
>
> > But in this case, I think the legal system agrees with me.
>
> In this case, I think you haven't _even_ disambiguated states of
> unhappiness from tort law (civil wrongs_.  And your problems go on from
> there.



Unhappiness can lead to forks, loss of userbase, and hence loss of revenue
for important contributors.  There is such a thing as general morale and
that is important to a project.

Note that I am one of a number of project leaders in one project.  Until I
hear otherwise, I am going to assume that Rick is an activist and advocate,
but not someone who has hands-on experience with leading an open source
project with a significant community.

> Ethics is only simple as long as people agree on what is moral.
>
> Tell me something I _don't_ know, for a change.  And, preferably,
> something factually correct, while you're at it.


I actually disagree with both of you (and I think most philosophy types
would also since Plato and Aristotle).  If we can agree on what is "good,"
then we can talk about ethics.  Morality is ethics plus some additional
aspects I would classify as emotional or perhaps aesthetic at best.

Best Wishes,
Chris Travers
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opensource.org/pipermail/license-discuss_lists.opensource.org/attachments/20070910/c2488f1d/attachment.html>


More information about the License-discuss mailing list