Tue Aug 30 17:20:58 CDT 2005
that whether or it was allowed to implement a few function in
some DLL in a LGPL Wine by linking to a proprietary library
depended on my intent. I reject this absurd idea most violently.
However consider for a moment: What if he is right?
This means that with a LGPL:ed Wine:=20
If somebody that had some proprietary library and thought hey
with this library it should be quite easy to implement these
functions in Wine that prevented these cool application to
work. Some programming later... Great it works... Now I call
Lindows, Red Hat, SuSE (whomever that distribute Wine) and
tell them that I can license this library for $5 per end user
to you. Some time later... Great at least some them accepted
and the money keeps streaming in. Life is good.
Is his intent bad? Would you sue him for violating the LGPL?
If he publish his proprietary library function wrappers=20
under the LGPL. If he didn't? What about the Linux distributor?
Under what circumstances would you sue it?
Joe Hacker has bought a proprietary library to use
with his non Wine related hobby. One day he realizes that the
one of his hobby related also proprietary Windows application
he dual boots to since it never worked under Wine could be
made to work by implementing the missing Wine functions
with that library. He does so, it takes a weekend.
It works. He puts out the patch to Wine under LGPL on a hobby
related website for other that also owns the application and=20
the library. He is happy since dual booting is but a memory.
Life is good.
Is his intent bad? Will you sue him?
If some months or so later the maker of the protrietary library finds
the patch and contacts the Linux distribution like in the case above.
Will you sue the him, the proprietary library company or the=20
All this begs the questions:
1. Isn't it so, that as I claimed that the LGPL really doesn't =
2. Or alternatively, if it really does make what some of the people=20
above do illegal, do we REALLY want to use it?
> It's pretty obvious that neither the a gpl=20
> license nor a bsd
> license are optimal. It only take the observation of the lack=20
> on consensus
> to see it.
> All that I can really offer is this suggestion: end feuding=20
> and start a
> process to
> 1. Gather the requirements to the licensing of the wine project
> 2. Determine if any combination of existing licenses=A0can meet the
> 3. If no combination of existing licenses will suffice, write=20
> a license
> that does.
> Additionally, I advise that everyone abandon there preconceptions of
> licensing. There is no mandate that I=A0know of=A0to=A0license the=20
> whole project
> under the same license, only use one license for any part of=20
> the project,
> only use existing licenses, license any part of the project=20
> to different
> groups under the same license.
Quite true. However, take care with viral licenses,
they might spill over to the other parts all the same.
More information about the wine-devel