Re: module license taints kernel.

David Schwartz <davids@xxxxxxxxxxxxx> writes:
Rainer Weikusat wrote:
David Schwartz <davids@xxxxxxxxxxxxx> writes:
That's exactly what I'm arguing. That a linker is just like
tar+gzip.

This statement has not become more true in the meantime. A linker is
not an archiver because it does not create archive files. An archiver
is not a linker because it does not create combined executables.

You can claim to have refuted any argument of the type "X is just like
Y" by showing that there is some difference between X and Y.

To the contrary, you try to make an argument of type 'X is just like
Y' by showing that there is a superficial similarity between the two.
Something like 'if we generously ignore the differences, the topics
are actually fairly similar'. Quel surprise.

However,to have actually refuted it, you have to show that the
difference *matters* in the context.

Both linkers and archivers "mechanically" combine two works without
any creativity. The result is the same as the input -- the two
original works aggregated.

This sentence is in itself nonsensical, because an 'aggregation of two
things' is obviously different from 'two separate things'. If it
wasn't, why would a program be needed to create it?

For copyright purposes, the distinction is creative combination that
can create a new derivative work. Neither linkers nor archivers can do
this. For copyright purposes, they are the same process.

This is as undisputed as 'the sun rises in the early morning' and as
relevant 'for copyright purposes' (which you yourself
state). Consequently, this perspective can be dropped entirely from
the discussions.

.