With the beginning of the second month, we had a discussion on choosing the next possible step. Possible next steps were: using baseexport to develop human readable LaTeX/Markdown export or refactor current nmlexport package. And we decided to proceed with NeuroML exporter and worked on thar for couple of weeks. The changes though were large but they were very straightforward because, in the previous package, the most part involved getting information from Brian objects. But now, with Baseexport’s dictionary we only need to paste the required information to the neuroml2/lems template. The previous version of the package was also well-documented and clear (thanks to the developer who was one of the previous students of GSoC with Brian) and thus refactoring was an easy job.