Matt Thompson

And that's one reason I've been using starship. I work on clusters where the bash is still in v3 land and the zsh isn't even an allowed login shell. Thankfully...

@czender So I tried that and it turns out if you remove a variable, it also works just fine. It's like whatever process `ncks` does to remove a variable "fixes"...

@czender Glad I'm not crazy! And as I said, if you just "make a copy" with: ``` ncks -h -v longitude,latitude,ref_2996.00,surf_ref_I_2996.00,I_2996.00 trial.nc4 foo.nc4 ``` it seems to work with that!...

I just hit this myself in test building pFUnit on an M1 Mac with a hand-built GCC. To wit, I found I needed for pFUnit v4.3.0: ```diff diff --git a/src/funit/fhamcrest/BaseDescription.F90...

@samcmill Interesting. Okay. If I have time, maybe I'll try and make some `intel_oneapi` and `intel_oneapi_runtime` building blocks based on `packages`. I just know in the past it was so...

Note that my home laptop where I tried this is a fun melange of randomness. If `nsh` somehow sees my zsh configuration, I can see all sorts of issues possible....

I was about to file an issue about this as I was reading about and experimenting with `cache_preemption` (see #206) and based my test on `f90tst_parallel3.F90`. (As we are trying...

Has there been any updates on this issue? I need to know whether I should start re-coding many of my build systems to allow for the change in `nf-config`. ETA:...

@edwardhartnett @WardF Thanks for looking at it. I'm fairly certain reverting the `NC_FLIBS` in `configure.ac` would do it, but maybe there was an important reason for changing it.

@edwardhartnett @WardF Has there been any discussion on this? I'm wondering if I'll need to fix many of my GNU Make, etc systems to reflect the change in `nf-config` behavior....