[Michael Tuexen <Michael.Tuexen@xxxxxxxxxxxxxxxxx>]

Dear all,

I do not understand the issue completely. I do know that some
companies have IPRs on protocols. But a protocol consists of
packet formats and procedures. The IPRs I know of focus NOT
on the packet format, but on the procedures and functionality
the protocol provides. It is not very good to have a patent on
a packet format, because then I can easily circumvent the patent
but simply modifying the packet format.

Why is this relevant for the discussion here? A basic dissector
only deals with the packet format of a protocol, it does not
know anything regarding the procedures of the protocol. Only if
you provide some analysis of what is going wrong (the detection
of TCP retransmissions, for example) you 'put' some knowledge
of the procedure on the dissector.

So I think (I'm NOT a lawyer) it should be not that problematic
for a company to provide a dissector for a protocol that might
be covered by some IPR. Simply make sure when writing these patents
not to patent the packet format.

However this does not solve the problem with Andreas H.323 stuff
which I think would be great to be included in ethereal.

From the discussion on the list I have the impression the there is
no agreement yet that protocols covered by IPR should be supported
but there IS a agreement that it would be great that Andreas H.323
stuff could be included in an ethereal distribution.

Best regards
Michael

On Friday, Nov 22, 2002, at 15:47 Europe/Berlin, Mark H. Wood wrote:

> On Thu, 21 Nov 2002, Christopher K.  St.  John wrote:
> > On Thu, 21 Nov 2002, ddutt@xxxxxxxxx wrote:
> > > What drawbacks do you see in changing the plugin license ? I haven't
> > > been able to make out what your worries are. Can you repeat them to
> > > me one more time ?
> >  The main problem is that the whole "The GPL stole my patent!"
> > argument is complete nonsense.
>
> I think you are correct to say that GPL doesn't invalidate a patent or
> make it unenforceable.  But that is not the point.
>
> The GPL requires that "any patent must be licensed for everyone's free 
> use
> or not licensed at all."  Well, what does "use" mean here?  Since 
> patents
> protect *ideas* rather than *expressions*, does this mean that in
> accepting a GPLed program I automatically receive a license to use the
> patented ideas embodied by it in *any* way, or only in the context of 
> the
> original work and any modification thereof which I might make?  *That*
> seems to be what makes the lawyers nervous.  Why would anyone pay for
> licenses when he could get them for free by accepting some bits of
> software, extracting the ideas, and throwing the software away?
>
> Unless clear limits are set on "use", there's no point in patenting 
> ideas
> embodied in GPLed code, because (since it may be that everybody thereby
> receives an unlimited license to practice those ideas) there is no
> practical way to maintain a monopoly in selling works based on your 
> ideas
> or to demand license fees of others who want to sell them -- which is 
> the
> whole reason for seeking patents.  As I reread the GPL, I become more
> firmly convinced that GPLing a work based on patented ideas should not 
> be
> done unless one has decided to stop using the relevant patent as a
> bargaining tool in other contexts.
>
> On the other hand, GPL also states that it applies to "any derivative 
> work
> *under copyright law*."  [Emphasis added.] I don't know precisely how
> copyright law defines "derivative work", so there may be limits on just
> how far you can remove the ideas from the code, beyond which GPL would 
> not
> apply.  Since the bit about patents "licensed for free use" occurs in 
> the
> Preamble and not in the Terms and Conditions, one could argue that any
> such limits supersede the possible blanket claim stated earlier.  (Does
> your head hurt yet?  Mine does.)
>
> Even if the "free use" bit were found to be limited to the original
> expression and reasonable modifications of it, I think that a GPLed
> project should not accept GPLed code containing patented ideas unless 
> an
> additional license is offered, granting permission to use the patented
> ideas specifically within such expressions.  Copyright and patent are
> related, but it is the *interaction* of the two which gives us trouble
> here and we need to deal with them separately in order to precisely
> define that relationship.
>
> So it does seem to me that either changing the license on the affected
> code, or establishing an "arms'-length" relationship between Ethereal 
> and
> any extensions containing patented ideas, would be a really good idea 
> if
> one is unwilling to turn down the offered extensions.
>
> (Disclosure time:  I just recalled that I own (very little) Cisco 
> stock.
> I believe that that does not influence my thinking on this matter, but
> you may choose to believe otherwise.)
>
> --
> Mark H. Wood, Lead System Programmer   mwood@xxxxxxxxx
> I continue to not be a lawyer.
>
> _______________________________________________
> Ethereal-dev mailing list
> Ethereal-dev@xxxxxxxxxxxx
> http://www.ethereal.com/mailman/listinfo/ethereal-dev