Makefile changes

[matt-pharr]

Changed makefiles to consolidate environment variables for different compilers in DEFAULTS.in, and fixed straggler REAL*8 -> REAL(r8) instances

[matt-pharr]

I did my best here to find what flags pgfortran would use but I have never used pgfortran before -- it seems like this was just used for GPU stuff related to #116.

[matt-pharr]

OK - this should be ready to merge. Note, the one non-logistical change I made here is c7e0b2b. I simply activated all three coil sets in the DIIID ideal example so that it was more directly comparable to the resistive/kinetic examples (and the more coils the merrier?). If it is preferable, we could also go the other way and make the resistive/kinetic examples only use the C coils.

[logan-nc]

hmmmm... As I mentioned in @matt-pharr 's commit c7e0b2b, the examples were previously used for regression by plotting outputs for various releases. I definitely see the benefit of having consistency between the 3 DIII-D examples though, as it allows a quick comparison of the responses for tutorials/learning/checking/etc.

@matt-pharr has improved the ability to script up checking out versions and running them, so I suppose plotting historical results can now be done that way. Is there something @matt-pharr can add to the source code that would allow others like myself to easily run such a test? Lets say, for example, that I would want to compare DCON dW for a given example case for all releases.

[logan-nc]

I'm currently thinking we should change the ideal example now, and try to do #208. Sound ok Matt?

[matt-pharr]

Ah -- following my comment on #208, I think I understand what you are looking for now.

Yeah, something like that can definitely be done. So you want to keep 6c4aa30 and c7e0b2b on this pull request and then do #208 on a separate one?

[logan-nc]

Yes. I merged this, counting on #208 materializing in the future.
Thanks!