[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Bug#189164: libdbd-mysql-perl uses GPL lib, may be used by GPL-incompatible apps

On Thu, May 29, 2003 at 11:57:18AM -0400, Nathanael Nerode wrote:
> Jakob Bohm said:
> >Anyway, I thought the common GPL linking claim was that the
> >runtime in-memory process image includes a copy of the GPL code
> >and is thus a derivative of that copy.
> But this "derivative" (if you assume that it is a derived work) is 
> created by the user on their home computer, and is not redistributed. 
> Since it's not "distributed or published", it's not subject to the GPL's 
> 'viral' requirements under section 2b:
> >b)  You must cause any work that you distribute or publish, that in
>                                       ^^^^^^^^^^^^^^^^^^^^^
> >whole or in part contains or is derived from the Program or any part 
> >thereof, to be licensed as a whole at no charge to all third parties 
> >under the terms of this License.
> (Emphasis mine)
> This is why the runtime in-memory state of the dynamically linked code 
> is not directly relevant to the GPL's 'virality'.


As I said, I believe there is a continuum of cases ranging from
mere aggregation to manually mixing of code lines in the
hand-edited source.

One point in this continuum is when a library and an
application are both in a Debian CD image, where the application
is clearly marked to automatically link to the library whenever
run.  Is this mere aggregation or do the two form a single work
derived (by simple inclusion) from the library.

However the main point of my post was not that.  My main point
was that in Borland vs. Lotus, the issue placed before the court
was the right to *re-implement* a compatible interface, not the
right to implement things that *use* the interface.

More specifically, this is how I recall the case (I was not
around, I just read about it once, so my memory may fail me):

Lotus had a monopoly status in the spreadsheet market with its
Lotus 1-2-3 program.  Part of its monopoly lock-in was the fact
that users had invested much time and effort creating and
sharing macros written in the Lotus macro languages.  Macros
which often encoded complex things like price schedules or U.S.
Tax Legislation, and were sold as commercial products or used as
necessary parts of business relationships (similar to the
current status of MS Word .doc files).  The Lotus 1-2-3 macro
language happened to consist of simply spelling out the actual
menu shortcut keys for the Lotus commands.

Borland, in its Quattro Spreadsheet program chose as a major
feature to implement full support for running and manipulating
existing Lotus macros, just as the Free Software community do
today for many proprietary interfaces.

Lotus sued, claiming that Borland broke the Lotus copyright on
the menu tree design detail and the artistic and random choices
amongst alternate naming and placement of the items.  Borland
said that such a claim would amount to forcing gratuitous
incompatibilities to the detriment of the general publics
ability to use different products at their choice.  A comparison
was made (in court) to Ford (hypothetically) demanding that all
non-Ford cars placed the pedals or steering wheel differently.

Borland won, re-implementing an old interface as part of a new
superior product while preserving the verbs of the command
language seem to be legal and not a violation of the copyright
of prior implementations.  The ruling also said that such
protection could be obtained with a patent, but Lotus did not
have one.  All of this in the U.S. of cause.


This message is hastily written, please ignore any unpleasant wordings,
do not consider it a binding commitment, even if its phrasing may
indicate so. Its contents may be deliberately or accidentally untrue.
Trademarks and other things belong to their owners, if any.

Reply to: