Hi Olaf,

Sorry it's taken me so long to get this to you. I wanted to do a bit more work
on it but I haven't had time. There are two directories here:

test1:  In this case, I used the more recent version of PyLith to generate the
        Green's functions. This version of PyLith allows you to use a basis order
        of 0 for the applied slip, which means you don't have to do your own
        integrations. There are two sets of GF: one for updip slip and the other
        for left-lateral slip. In the output, the first component of both slip
        and traction corresponds to updip, the second component corresponds to
        left-lateral, and the third component corresponds to normal. For these
        files both the slip and the resulting traction are in the same file.

test2:  In this case I used PyLith v2.x to generate the Green's functions. I computed
        them on a much finer mesh and then integrated to get the results for a coarser
        mesh. Note that this mesh is not really suited for simulators (I was using it
        for geodetic inversions). There are multiple files and in this case I use known
        convergence directions (the par and perp files). For these files the applied
        slip amounts are in the *impulse* files and the resulting tractions are in the
        *tractions_traction* files.

I need to do a lot more work to make sure I've done things correctly, and I also want
to see if using a basis order of 0 for fault slip gives us sufficient resolution of the
seismic velocity model. Anyway, hopefully this is enough to get you started and please
email me if you have any questions.