Import of XZ (updated)

[Joerg Sonnenberger <joerg%britannica.bec.de@localhost>]

Hi all,
I've been asked to summarize the discussion so far and reflect that in
the proposal.

I want to import the non-GPL3 parts of XZ. Integration into the base
system will do done in a way that ensures that machines with a
reasonable amount of memory can decompress the results without starting
to swap. In short, this means that the default usage will be the
equivalent of xz -6 or some other limit to decide separately. The
concern about files created with non-default larger dictionary sizes has
been addressed upstream already by dropping the check for the decoder.

Joerg