This is not just a GPL issue [LWN.net]

This is not just a GPL issue

Posted Jan 6, 2003 0:00 UTC (Mon) by mbp (guest, #2737)
In reply to: This is not just a GPL issue by ksmathers
Parent article: On the licensing of software patents

The point is that it may make it impossible, or difficult, to distribute and use GPL'd software in practice.

Software patents are a terrible problem for the industry as a whole, and for free software in particular. (See for example Phil Karn's comments: <http://people.qualcomm.com/karn/patents/patents.html>) A supposed open standards body has no business encouraging their further use, or their incorporation into standards.

(Log in to post comments)

This is not just a GPL issue

Posted Jan 16, 2003 14:24 UTC (Thu) by jdthood (guest, #4157) [Link]

Not so.

Let's invent an example to make this clearer. Suppose a web standard is introduced that involves a patented algorithm for encrypting information in transit. Company C owns the patent, but agrees to license the patent royalty-free for use on the web.

Hacker H1 writes a program that implements the algorithm in a browser plugin and publishes it under the GPL.

User U1 uses this plugin to access a secure website.

So far, no one has broken the law. U1 used the algorithm within the terms of the patent.

Now, Hacker H2 modifies H1's program to encrypt traffic between a database client and server. He releases his program under the GPL.

U2 uses H2's program to access databases securely.

I suspect that either H2 or U2 has broken the law. However, that does not change the fact that the original article got it wrong. LWN wrote:
    ... if a W3C standard contains patented
    technology with "field of use" restrictions,
    no implementation of that standard may be
    distributed under the GPL. The GPL does not
    allow that sort of restrictions.
That is wrong. Such software *can* be distributed. There is no incompatibility between the GPL and field- of-use restrictions. The GPL disallows the introduction of restrictions *into the copyright license*, but says nothing about restrictions which may exist independently of that license.

This is important. If I am write then I hope that LWN.NET will publish a correction at some point. Otherwise, someone please correct me.
--
Thomas Hood

This is not just a GPL issue

Posted Jan 16, 2003 14:39 UTC (Thu) by corbet (editor, #1) [Link]

"The GPL disallows the introduction of restrictions *into the copyright license*, but says nothing about restrictions which may exist independently of that license."

What I wrote is based on section 7 of the GPL:

If, as a consequence of a court judgment or allegation of patent infringement or for any other reason (not limited to patent issues), conditions are imposed on you (whether by court order, agreement or otherwise) that contradict the conditions of this License, they do not excuse you from the conditions of this License. If you cannot distribute so as to satisfy simultaneously your obligations under this License and any other pertinent obligations, then as a consequence you may not distribute the Program at all. For example, if a patent license would not permit royalty-free redistribution of the Program by all those who receive copies directly or indirectly through you, then the only way you could satisfy both it and this License would be to refrain entirely from distribution of the Program.

The GPL disallows restrictions on derived works. Field of use restrictions are a restriction on derived works. I believe that what I wrote is correct.

This is not just a GPL issue

Posted Jan 16, 2003 22:24 UTC (Thu) by jdthood (guest, #4157) [Link]

Hmm.

> For example, if a patent license would not permit
> royalty-free redistribution of the Program by all
> those who receive copies directly or indirectly
> through you, then the only way you could satisfy
> both it and this License would be to refrain
> entirely from distribution of the Program.

You are right that the GPL does say something about
restrictions that exist independently of the license.

But I maintain my disagreement ... for now. The way I read it,
section 7 is redundant or nearly so: it just emphasizes
that if you can't legally (re)distribute the work under
the GPL then you can't (re)distribute it at all. But
this already follows from the rest of the license. I
suppose the section was added in order to forestall
(by making its illegality more explicit) an obvious dodging
tactic.

The crucial question is: DO the conditions imposed by
field-of-use patents somehow "contradict the conditions of
this License"? If so, then the work cannot be distributed
under the GPL. If not then it can. Let us try to answer
this question.

You say that "the GPL disallows restrictions on derived works"
and reason that because a field-of-use patent imposes
restrictions, implementations of such patented ideas cannot
be distributed under the GPL.

However, I think that there is a slip in the reasoning here.
The GPL disallows the introduction of more restrictive terms
into the copyright license of the derived work to be
distributed. A field-of-use patent imposes restrictions quite
independently of the copyright license. Its authority
derives from an entirely different law. Obviously the GPL
cannot disallow *these* sorts of restrictions. If it did,
then nothing at all could ever be published under the
GPL, because all software distribution is restricted by some
part of the law. The law says (I presume) that I cannot
distribute software that contains child pornography. This
is a restriction, but surely it is not a restriction that
conflicts with the GPL.

Regards,
--
Thomas Hood

This is not just a GPL issue

Posted Jan 16, 2003 22:45 UTC (Thu) by jdthood (guest, #4157) [Link]

Just thought it might be useful to add a link to a posting
by Bruce Perens on this subject in debian-devel mailing list.

http://marc.theaimsgroup.com/?l=debian-user&m=104138016327196&w=2

He doesn't see a problem with GPL implementations of
field-of-use patent encumbered algorithms, so long as
the relevant code is not used to do anything outside
the permitted field.