Open moorepants opened 8 years ago
ashutoshsaboo
commented
8 years ago :+1: I completely agree with this @moorepants . We must definitely have research examples.
What we can do is, write at-least 1 example for each algorithm that we write in the sympy-paper. And in addition to this, we can write some examples (3-4 or maybe more) for the entire features section in the paper, to highlight the more important features of SymPy through those examples.
moorepants
commented
8 years ago I think we'd first need to compile a list of research examples and see what features are used in those examples. We can do the same with industry and education. Once we have this list of features that are used by these domains, it will help us decide which feature should be highlighted in the paper (i.e. the ones that people actually use).
The scikit-image paper also has a nice section on comparing their software to similar libraries. We have a section on mathematica comparison, but could probably add more there too.
After skimming through the scikit-image paper (https://peerj.com/articles/453/) again I really liked the "Usage Examples" section that shows how the software is used in research, education, and industry. This is a very important thing to tell a story about. It's one thing to show that we have this software and these are the features, but being able to show how the software is powering these three topics will make the paper very strong. I would like to advocate that we restructure things some to make this a prominent part of the sympy paper.