# Week 10 Feedback notes

- Some people were doing coursework but most were engaged, people seemed to find the idea of using a super-computer interesting.
- Most people managed to get at least the HelloWorld and some of the pi example running.
- Students spent a lot of time waiting around for submitted jobs to run, so we can maybe focus more on how to get a good workflow of local testing and then submitting jobs
- We could maybe cut the serial job example as it's not any simpler to submit and run than the mpi example and it delays the use of MPI
- Many issues with things like copy and paste into ssh so things would run smoother if we just add the job scripts to the git.
- A summary of qstat status "Eqw" "qw" "r" etc. and output types (.e, .o, .pe, .po) would be useful to have with the exercises so people can check what is going on.
- Most people didn't get to the ping-pong example so we might want to speed up getting to the actual MPI coding parts a little bit. Session focus felt a bit more on myriad / job submissions than distributed programming.
- Need to settle on C bindings (mpicc as in exercises) or C++ bindings (mpic++ as in notes) for consistency