Reviewer kanishkan91's comments for JOSS paper #154

Comments from kanishkan91 copied from this page:

Just finished my review. This took much longer than expected, however, I am happy to have reviewed this package. Really useful package that demonstrates clear utility (especially for adjacent earth system modellers such as myself). However, I had certain comments. I don't want to open GitHub issues for these since these have nothing to do with the functioning of the package but rather with the package architecture.

1. Compatilibility with E3SM outputs vs CESM/CLM outputs- While I understand that this app is designed for E3SM specifically , how flexible would this app be to other similar earth system models such as CESM which E3SM has inherited a lot of its functions from? I wanted to raise this since this speaks to the "general usability" of a package such as this. This should also be explicitly discussed in the text.

To do:

• Improve the gallery page using cards and filter (PR docs: reorganize gallery using cards and filter #156).
• Address issue Website: add example and mention of CAM's pg3 grid #152 to provide an example for CESM output.
• Address issue Website: add gallery examples for dycore's np4 grid #153 to provide an example for EAM dycore output.
• Mention these, as well as ELM examples in gallery, in the paper.

2. Differences from ParaView- While this app uses ParaView (as noted in the paper), I believe more details are required regarding how the app functioning, architecture departs from ParaVieew. For example, if there were significant updates to ParaView, would that hinder the functining of your package? This should also be discussed in your repo's README and in the paper.

3. ParaView installation- I must confess that I had to spend some time installing ParaView on my machine. There seem to be a lot of discussions regarding troubleshooting its installation on several forums. Could you add some standardized instructions on how to install this for a new user? Again, this may be something that is needed in the README or docs rather than the paper.

4. Statement regarding the memory required on local machines- The utilization of the developer version of this package seems to require a lot of free memory to download ParaView, connectivity files and simulation files. This would necessitate a statement from the authors regarding the amount of free space that they recommend users start with.

5. Automated tests- Revisiting @kdorheim's question above, what automated tests have been implemented for this app. I understand that since this is an app, writing unit tests in the general sense may not be applicable. Yet just some simple tests to make sure that this launches or produces outputs would be useful.

Those were my major comments. If you can respond to them and kdorheim's comments over the next couple weeks, we can take it from there. Again, sorry about the delay.