onboarding ideas about software-review HOT 10 CLOSED

Not from the unconf, but something to elevate to updated guidance for authors, from the review of rnaturalearth:

If you feel your reviewers have substantially contributed to the development of your package, we suggest you credit them in the DESCRIPTION file Authors: field as reviewers, like so:

person("Lincoln", "Mullen", role = c("rev"))

from software-review.

Seems like a good idea - can you add to guide for authors

from software-review.

Done. Also, new life goals for how I am listed in a future package:

> utils:::MARC_relator_db[utils:::MARC_relator_db$code %in% c('wde', 'wdc'),]
             term code                                                                                                description usage
260 Wood engraver  wde  A person or organization who makes prints by cutting the image in relief on the end-grain of a wood block      
261    Woodcutter  wdc A person or organization who makes prints by cutting the image in relief on the plank side of a wood block 

from software-review.

I think we want a formatted report rather than just logs, with things ordered from most important (editor-rejects) to 'useful to know' info for reviewers.

Other things to potentially run in the the check/put in the report:

• Just the NOTES/WARNINGS/ERRORS from R CMD Check
• exported functions use the same naming style
• what CI service is being used? (A Linux AND Windows CI?)
• README.Rmd is used if there is substantial code in README
• Lines of code in .R files, Roxygen comments, and tests (I think measuring this and a little more experience might be useful before we move to limiting review size).
• List of packages in Depends
• Is there a code of conduct?
• print() or cat() calls in functions

from software-review.

I think I'm going to start a discuss thread for reviewers to ask them what would be useful in an automated report.

from software-review.

Haha that's awesome - your code confirms that my guess of 'rev' was about right - not sure why R packages need wood engravers though - maybe to ensure their longevity after the internet goes down? ;)

utils:::MARC_relator_db[utils:::MARC_relator_db$code %in% c('rev', 'wdc'),]
          term code                                                                                                description
223   Reviewer  rev           A person or organization responsible for the review of a book, motion picture, performance, etc.
261 Woodcutter  wdc A person or organization who makes prints by cutting the image in relief on the plank side of a wood block
    usage

Many thanks @noamross for showing me what those codes mean - any more info where those codes come from appreciated.

from software-review.

start a discuss thread

okay

from software-review.

@Robinlovelace A little googling reveals that these are codes for MARC bibliographic standards, which are what Library of Congress uses for metadata: http://www.loc.gov/marc/ , https://www.loc.gov/marc/relators/relaterm.html

from software-review.

Thanks for the info - impressed R incoroporates those codes into person() - apologies for being slow, see now that's also linked to in ?person.

from software-review.

discussion at https://discuss.ropensci.org/t/what-additional-automated-checks-would-be-useful-for-package-review/347

from software-review.