You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The thing is, I'd also like an mpiuni case but I can't expose a netcdf built with MPI to an mpiuni case. So, at the moment I've done the following where my mpiuni is now using a different compiler:
But I'd like to move those ugly modulefiles to a less public directory just because I don't want people seeing them. But that means I need to add a new module_path and from my read of the python, module_path (at the top for a machine) does not expect a list.
@theurich or other gurus,
I recently tried to add support for netcdf in my discover tests via an ugly modulefile which exposes a Baselibs build of netcdf. As in:
The thing is, I'd also like an mpiuni case but I can't expose a netcdf built with MPI to an
mpiuni
case. So, at the moment I've done the following where mympiuni
is now using a different compiler:But I'd like to move those ugly modulefiles to a less public directory just because I don't want people seeing them. But that means I need to add a new
module_path
and from my read of the python,module_path
(at the top for a machine) does not expect a list.So I wonder, could this work:
That is, I don't know how deep you can put some of these keys.
The text was updated successfully, but these errors were encountered: