Google is Violating LGPL Source Code

[Daniel Berlin]

On Tue, Apr 7, 2009 at 12:34 PM, Daniel Berlin <dannyb at google.com> wrote:
> On Tue, Apr 7, 2009 at 11:58 AM, Stephen C. Fedder <sfedder at hotmail.com> wrote:
>> Pardon me if I am out of line following this thread, but Google is already releasing GWT completely as open-source under Apache 2.0 license, so it is not as if they are not releasing source code for the product in question.  Also it seems as if this whole thread would have not been necessary if the original writer would have gone to the extent of actually verifying that the LGPL package was in fact actually being distributed by examining the downloads before publicly accusing a company like Google which contributes a huge amount of software and resources as open-source of violating license.  Even if a company appears to be in violation, it is always the best course to first pursue compliance with them directly rather than posting public accusations.  It looks like the biggest thing that Google has done here is to have an erroneous, parhaps out of date, reference in some of their online documentation referring to LGPL packages that may be used with GWT by the end user being provided by Google.  I'm pretty sure there is no legal action to be taken as a result of a mistake in documentation.
>>
> One of the GWT developers now tells me the jfreechart stuff is
> actually in there, just buried fairly deep (it's in a .war file
> contained in a .jar file somewhere) these days.
> So I apparently misspoke on that point :)
> That said, ...
>
>> I have seen plenty of threads before regarding use of GPL/LGPL in devices where the accuser is told that they need a distribution and positive proof, i.e. dumps > or listings showing the actual presence of GPL/LGPL code in a distributed product before posting an accusation.  Stating that companies who actually do
>> distribute LGPL binaries also need to post or otherwise make available corresponding source is just restating the license which anyone who follows this list
>> should already know.
>
> So Mr Hardy's argument is essentially (as far as i understand it,
> please correct me if i am wrong):
>
> Distribution of gwt with jfreechart jar files embedded in it is not a
> "work that uses a library", instead gwt is the "work that uses the
> library" and it happens to include jfreechart.  IE he doesn't believe
> the combined work is the work the LGPL is talking about in section 6
> Thus, we must comply with both section 6 (for the GWT work), and a
> bunch of other sections for the jfreechart work.

BTW, just to split this thread,, my reason for disagreement with him
on this point simple. The LGPL (2.1) states "6. As an exception to the
Sections above, you may also combine or link a "work that uses the
Library" with the Library to produce a work containing portions of the
Library, and distribute that work  ...."

It clearly says "distribute *that* work" (emphasis mine) where "that"
is referring to the combination of the LGPL'd library (jfreechart) and
the "work that uses the library" (gwt).
Thus, the work they are referring to as being granted exceptions to
the other sections above is the combined work of GWT with jfreechart
(whose combining "produce[d] a work containing portions of the
Library").
He disagrees with this, and believes the inclusion of jfreechart jars
in the work is not covered here.
As I said however, whether I agree with him or not, we are still
(AFAICT) in compliance in either case.

--Dan