I like the fact that checklist errors always explain clearly what has to be done, and mostly this works fine. But in the few cases where it doesn't (often due to a different behaviour in different OS), I think it is a drawback that even in the logging it is unclear what causes the error. Would it be a possibility that for instance in this case, the differences between the NAMESPACE files, or even just both versions of the file, could be added to the logging? And more generally: could some more outputs be generated in the logging for the checklist specific tests?
[Here](https://github.com/inbo/inbodb/pull/35#:~:text=check%20package%3A%20locally%20I%20don%27t%20get%20the%20error%20that%20suggests%20to%20run%20devtools%3A%3Adocument().%20Running%20devtools%3A%3Adocument()%20locally%20(with%20roxygen2%207.1.2%20installed)%20rewrites%20the%20NAMESPACE%20file%2C%20but%20apparently%20without%20any%20changes%2C%20as%20the%20file%20does%20not%20appear%20in%20the%20git%20window%20(also%20not%20after%20refreshing)) I describe an error that is unclear to me: it suggests to run
devtools::document()
and push an updated version ofNAMESPACE
, but running this command locally doesn't generate a new version of theNAMESPACE
file. And the logging doesn't give the slightest hint on what has been tested, and what could be wrong there.I like the fact that checklist errors always explain clearly what has to be done, and mostly this works fine. But in the few cases where it doesn't (often due to a different behaviour in different OS), I think it is a drawback that even in the logging it is unclear what causes the error. Would it be a possibility that for instance in this case, the differences between the
NAMESPACE
files, or even just both versions of the file, could be added to the logging? And more generally: could some more outputs be generated in the logging for the checklist specific tests?