[Qt-interest] LGPL and static linking

[David Ching]

"Stephen Jackson" <spjackson42 at gmail.com> wrote in message 
news:43fef2450911251525y4946345as9f1e9e8109031ffb at mail.gmail.com...
> If that is true then I am astounded. If Nokia is concerned about
> "vagueness of the terms in the LPGL 2.1 wording", then why ever did it
> choose to adopt this licence rather than some other one that says
> precisely what Nokia intends to say?
>

I asked that, the answer was Nokia did not think the world needed yet 
another LPGL-like license that would take time to digest and understand. 
They preferred the well-known LPGL 2.1 as that has existed for some time, 
and people (those who can afford lawyers anyway) have already done their own 
investigations and know what it means for them.  And this was from the woman 
who wrote the Qt Commercial license, so clearly Nokia is not shy of writing 
their own license when they believe it in their interest.

She said they purposely did not choose the latest LPGL 3.0 license as that 
has even more room for different interpretations due to the FSF's insistence 
on using non-legal terms that do not have clear meaning in any jurisdiction 
(as opposed to LPGL 2.1 having clear meanings but potentially different 
interpretations in different jurisdictions).

Regarding the different interpretations about whether static linking is 
permitted under LPGL 2.1, I believe she said the specific interpretation at 
issue is whether static linking a library with your own code causes your app 
to be interpreted as a derivative of the library it is static linked to (in 
which case you would need to publish your app's source code).  Or if static 
linking leaves your app merely a user of the library it is statically linked 
to, and thus does not require you to publish your app's source code.
This question does not arise if you dynamically link because your app's code 
is rendered into a separate physical entity than the library it is linked 
to, so it is clearly seen as a user of the library.

-- David