[Wien] follow up - *warning* ENE:

[Stefaan Cottenier]

> i don't feel i can rely on the total energy if there is a "*warning*" next
to ENE:.

May be, may be not. You should test this. Probably you want to compare
cohesive energies with and without the vacancy? Then calculate 1) your pure
material, and 2) a very small supercell (2x1x1 ?) with a vacancy. Do this
for several RKmax (5.5 -> 7.5 ?), and compare the differences in cohesive
energies. If you are lucky, you maybe have good results already for
Rkmax=6.0, which might be even larger than your reduced Rkmax (check the
latter by 'grep :RKM case.scf').

> In the same token i want a pretty good k-mesh for the vacancy
> calculation so i guess i have to sacrifice size of the supercell to do
> this....otherwise RKMAX becomes large for NMATMAX.  Is this correct?

The k-mesh doesn't affect Rkmax, only the calculation time. Again, you
should check the k-mesh first, by a similar procedure as given above.
Afterwards, you can do your big calculation immediately with a k-mesh you
know you can trust (and which is not unnecessarily large).

Stefaan