ietf-mxcomp
[Top] [All Lists]

DEPLOY: Microsoft Patent license unworkable with GPLed MTAs

2004-08-24 17:18:54

  Although I have been moderately active on the spf-discuss mailing
list, I have only been lurking on this list, but for some time now. 
Nevertheless, now that Microsoft has only just now, as of last night,
published its license for its IPR claims on unspecified technologies in
the -core and -pra drafts, I submit my evaluation of it and my concern
as MARID moves forward.
  Let me preface this by saying that being told to speak with my lawyer
to contact Microsoft's legal department is not, in my opinion, an answer
that is acceptable at this late stage.  It has taken Harry Katz at least
a month to go from saying that he will be working with Microsoft
attorneys to get an answer to last night's posting of the updated
license.  Why should a non-Microsoft employee expect to get a quicker
answer to get things clarified?  I thought one of the co-chairs made it
clear about month ago that when participating in this WG, we participate
as individuals.  Why should anyone on this list be directed to legal
department of the employer of someone on this list if that is, in fact,
the case?
  And on the issue of 'speaking to my lawyer', that is not something
many GPL licensed software authors can afford.  Even those who are small
businesses.  This list was promised clarification and this list is where
any clarifications should take place.  And no, I do not believe that has
been accomplished.
  Next, I'm boggled by Eric Allman first saying that it isn't
"completely clear" that the Defensive Suspension and
non-sublicenseability clauses interact with the GPL and then going on to
conclude that Sender-ID should be endorsed, anyhow.  And saying that
Eben Moglen found such license conditions acceptable is a stretch.  (And
as a side note, and an honest question, were these Office XML patents
being submitted as part of an IETF standard?  If not, then it is
irrelevant to this discussion.)  In that same quote paragraph, Mr.
Moglen says 'This is not a license I would like to accept', but overall,
in the area of Office XML patents, he also pointed out that it was
unclear *what* patents Microsoft had.  Later in that article, Mr. Moglen
explains that:

        "Section 7 of the GPL says if you're restricted by a judgement
        or a license you've accepted, and they're incompatible, then you
        can't distribute your software under GPL. But it's not the case
        that this license violates Section 7."

From my reading this indicates that if I sign a license with an IPR
holder for Sender-ID and it turns out that it conflicts with the GPL,
then I can't make my planned enhancements the Exim MTA and distribute
those changes.  It would violate the GPL since I cannot confer upon each
user that I've distributed my modified version of Exim to the same
rights that I had when I obtained my copy of Exim: to modify it and
distribute his modifications without restriction.
  As a email hosting provider and consulting company we were planning on
enhancing Exim for this purpose.  Though we have traditionally
implemented sendmail, we are always looking at our options, as we prefer
the GPL.  We cannot legally distribute our planned modification of Exim
if our customers do not have the same rights as us to distribute the
source code as it would violate the GPL.  Our resources are limited.  If
we must talk to a lawyer to determine if we are wrong and this is okay,
it would likely kill the project for us.
  Sure, I could risk it and just implement Sender-ID without signing a
license with any Sender-ID IPR holders and distribute my modified
version of Exim.  Then I would be in full compliance with the GPL (from
my reading of Mr. Moglen's statement above), but I'd be at risk of
cease-and-desist type legal action from the IPR holders.  I believe this
is exactly what authors of GPLed mp3 software are risking...that Thomson
Multimedia will not pursue them for royalties...a risk that Red Hat, for
one, is not willing to take since Thomson made some changes to its web
site.  Is this WG endorsing this kind of activity?  (That is, of course,
a rhetorical question.)
  I think one thing that is being misunderstood by some on this list is
that, with respect to the GPL, the end user is absolutely no different
than an intermediate developer.  By intermediate, I mean someone who
modifies and distributes his modified version of someone else's
copyrighted, GPL licensed, work.  I, as an end user, have the very same
rights, without entering into any written contractual agreement with any
third party.  This patent license eliminates that right I have as an end
user of a GPL licensed software package with the 'non-sublicenseable'
term alone in 2.1 and 2.2 among other places.
  This license needs more work to become GPL compatible.  The GPL is
nowhere near as complex as some would like to make it out to be.  And it
doesn't take a lawyer to see that the non-sublicenseability, Defensive
Suspension, Identification of Licenses and other sections are not
compatible with the GPL.
  Not to pick on Eric Allman too much, but I have to comment on his
opening of Sendmail, Inc.'s position as well:

        "Sendmail, Inc. has worked with Microsoft to help them produce a 
        patent license that will be acceptable for ourselves, the IETF, and 
        the open source community at large."

Unfortunately, though Sendmail and Microsoft may agree that this license
is acceptable to each other and possible even to the IETF, I think you
will find that it is not at all acceptable to the 'open source
community' at large, nor, (to be fair to the FSF) to the Free Software
movement.
  I realize that there are many who dislike the GPL and how difficult it
makes it for companies who wish continue in this patent cold-war.  It is
controversial to some, revolutionary to others (myself included), and
*blah* to others.  I am not posting to debate the relative merits of any
license vs. another or any business model vs. another.  That is
irrelevant.  What is relevant is that there are AT LEAST three different
MTAs covered by the GPL that will be ghetto-ized by this licenses (those
three being Exim, PowerMail, and Courier-MTA).  SMTP has been
implemented by GPL, BSD, proprietary and other-licensed software with no
encumberances for what, two decades, or nearly so?  Why should we
unfairly push GPLed MTAs out of the market now?  I believe we should
not.  Unequivocally.
  I think we need to have Eben Moglen or Larry Lessig look at this new
modified license and comment, rather than rely on some past comment made
by Eben Moglen on a *different* license concerning a technology that
wasn't on a standards track (that I know of, anyhow), i.e.: unrelated.
  But, as we are now in last call, I have to wonder if we really have
enough time for that.  And that is of NO fault to anyone now posting in
opposition to this license.  We didn't wait until the last minute...it
is Microsoft itself who waited until the last call to post its modified
license.  We were just waiting until the promised clarification was hand
to us.  There was no real way to respond until then.  And given that no
technologies have been declared with specificity over which Microsoft
claims it *might* have IPR, I recommend that Sender-ID not be submitted
at this time and instead SPF-Classic/SRS be submitted.
-- 
-Paul Iadonisi
 Senior System Administrator
 Red Hat Certified Engineer / Local Linux Lobbyist
 Ever see a penguin fly?  --  Try Linux.
 GPL all the way: Sell services, don't lease secrets


<Prev in Thread] Current Thread [Next in Thread>
  • DEPLOY: Microsoft Patent license unworkable with GPLed MTAs, Paul Iadonisi <=