epsilon.cplx.x complains jdftx wfn not normalized, off by about .02

[mike marchywka]

I'm using an isolated water molecule similar to the jdftx example on 

their website. I added a few more empty states and the gw dump code

but otherwise similar.

I took a look at the real space waverfunctions dumped separately to test

some code and they appear to vary by almost 10 percent in some cases

and my normalization agrees with epsilon's message. Presumably this

is an issue with jdftx or integration scheme. But for now I just wanted

to proceed with this value but can't find a keyword to accept it.

Are there issues with charge neutrality or something that come up?

Thanks.

[mike marchywka]

wrong pseudopotentials seems to run ok now...