VanLUG Email Archive
At the risk of starting this up again ;^), I'd like to ask a question --
I hope I can learn something here.
On Fri, 11 Sep 1998, Ted Powell wrote:
> On Fri, Sep 11, 1998 at 12:42:50PM -0700, Curt Sampson wrote:
> > On Fri, 11 Sep 1998, Ted Powell wrote:
> > > I suggest that if those hundred lines can't be sequestered in their own
> > > linkable/loadable/whatever module, then there's something _very_ special
> > > about them.
> > Not really, no. It's often not convenient, and sometimes not
> > possible, to have kernel code in an LKM, for example. In other
> > cases, you're dealing with a program that you would want to work
> > when statically linked (in single-user mode, for example), and you
> > don't want to bring in all the baggage of dynamic loading (the
> > routines for which would also have to be statically linked into
> > that program)
> I've just skimmed the General and the Library licences. It appears that
> if the code you want to use is under the General license, and you can't
> (or don't want to take the time to) get permission to use the hundred
> lines in question under the Library license, then you are indeed out of
> luck as far as incorporating the code into a _proprietary_ program is
> concerned. But if you want to incorporate it into a Berkeley-licensed
> program, section 10 of the GPL may apply.
All this being said, isn't it possible, if you want to use 100 lines
of GPL'ed code (or I should say, you want to do what those lines do),
to use a "clean room" procedure? You, the code designer, looks over
the 100 lines of GPL'ed code, writes and exact spec for code that
will do what that 100 lines does, then hand the spec over to another
coder who has never seen the GPL'ed code and who can write the equivalent
function for you?
Or is that too much to ask, or too much trouble? I honestly don't know,
I just work on Fortran code for my Ph.D. thesis :^/
Clemmitt Sigler
Va. Tech Physics Dept.