EGaraldi
commented
2 years ago Dear Ilian, please accept my apologies for the long silence. I had some personal issues in 2020 that forced me to focus away from this project. I then had to find a new job, and by the time this was done the JOSS submission was closed for inactivity, so I drifted away from this project. I have now finally restarted working on this project, so I want to honor your review work even though the submission is closed. Of course, you can expect a (much much) quicker response from now on.
I have fixed the tutorial mistakes and expanded and improved the usability of the code. I think you pinpointed some key questions about the future and success of this code. While I hope for an audience as broad as possible, which can take up part of the effort in keeping the database up to date, I believe it will realistically be picked up by a subset of the community. However, I do not have a particular audience in mind at this moment.
The CoRECon code is a useful collection of current reionization constraints from various collections of observational data. By its nature, it is intended as an expandable framework for including additional data when this becomes available.
The code was easy to install and to get started with. The update mechanism works well (though there was no new version of the code for me to test on).
However, the code can certainly do with some usability improvements. The one and only worked out example provided did not work as given in the online manual. First, the example code gave an error (which was easily fixed by commenting a line, but should nonetheless not happen). Second, while it did produce a plot as it was supposed to, that plot does not look anything like what it is expected to according to the manual. The test code uses completely different data sets from the example plot and is over different redshift interval. This is confusing, and not very useful as a test if the installation is working as expected. Some more, detailed (and properly working) examples are needed to get people started.
Apart from this, the various available functions are reasonably well (if briefly) documented. Adding additional constrains is also reasonably straightforward, and this is important given the rapidly evolving field. Somewhat less clear to me is if the author plans to regularly update the constraints. If this code gets picked up by the community these updates could be done by the community, but I am not sure how much this could be relied on. Without regular updates this code is quickly going to become obsolete. Most successful codes tend to have at least a few dedicated contributors.
One thing that was not quite clear to me is who are the intended users of this code. The people working in ths area for long time likely have some form of this already implemented, so might not be the target users. More likely this could be used by younger, less experienced researchers, like PhD students, or undergraduates. If that is the case, the code should probably be a bit more user-friendly than it currently is, to ensure this is picked up by a wider community, which in turn will make it more useful if additional data is continually being added.