Check compliance with ropensci guidelines

[adamobeng]

[adamobeng]

Here are some possible mismatches with the guidelines:

Packaging guidelines

• The order of the README sections is off
• We don't consistently use snake_case
• No rOpenSci footer, but that makes sense
• No code of conduct
• Missing git tags for CRAN releases
• We don't use httr for http requests
• We have some calls to print outside of print.*() methods

Wickham's "R packages"

• There are a lot of problems with our style as reported by both formatR and lintr

[kbenoit]

• We can update README
• I don't agree with the diktat about snake_case
• rOpenSci footer only comes if you submit to rOpenSci and are accepted
• Could add a code of conduct. It should include a prohibition on accessing object internals directly :-)
• We should add (and I think you have been adding) git tags for CRAN releases, including one I just submitted - and resubmitted - and resubmitted - yesterday and this morning. (If one offending line in the C++ now passes CHECK on Solaris, then it will stay and we can tag it.)
• Could and should switch to httr
• We don't have many (any?) calls to print() that are not designed to invoke print methods, but we do have a lot of cat() calls. It would be easy to hunt these down and replace with messages() calls.

On the formatR and lintR, I am pretty good about complying, but let's do a review of the results and consider what to clean up.

[adamobeng]

This branch is available to implement the changes. In that branch:

• I have updated the README
• I have replaced download.file with httr
• I am working on replacing calls to cat, but it's not a simple find-and-replace
• lintR has this output, and formatR makes these changes
• WONTFIX snake_case, rOpenSci footer

[adamobeng]

Re-opening final remaining issue as #234