william-dawson

I think we can also add the Fujitsu wrappers to the list of MPI wrappers. ``` diff --git a/src/tools/guess-mpidefs b/src/tools/guess-mpidefs index 2e9dfb670e..f553847bae 100755 --- a/src/tools/guess-mpidefs +++ b/src/tools/guess-mpidefs @@ -1,10 +1,10...

@jeffhammond It kind of works. I've gotten it to compile the target `nwdft`. I can run basic SCF calculations (with an hcore guess). But there are still bugs to fix,...

Thanks @yaoyi92 , yes the compilation does work with the master branch. I think releasing version 2.9.1 makes sense and would solve this issue.

@jan-janssen I see you have made various attempts at building a BigDFT package with conda. I think that this PR should take care of things. If you have any suggestions...

@chrisjsewell That is very interesting. With BigDFT, you will have the same problem as abinit in that it depends on libxc version 4. I suppose this is why the spack...

@jan-janssen I believe all the tasks are completed. Using `channel_targets` in `conda_build_config.yaml`, I've designated this as a release candidate. If everything looks good to you, I'd like to merge this...

> Is `libfmalloc` part of this package or are you redistributing? > > Shipping static libraries is not allowed. The following libs are being shipped statically: abinit, atlab, bigdft, CheSS,...

@carterbox I've gone through the most recent suggestions. For the testing of the dynamic libraries you suggested, I had to remove `liborbs` from the list, as there is an issue...

> Apparently, I'm the only reviewer of compiled packages right now. 😆 Be sure to add that you to CV 😄

One enhancement has been added to improve error control. Now, one can use the energy of the system as a measure of convergence when performing density matrix calculations.