Update DOI and LICENSE for CRAN submission

[hongyuanjia]

First submission try, still got a NOTE. CRAN maintainer replied:

Thanks, we see:

   License components with restrictions and base license permitting such:
     MIT + file LICENSE
   File 'LICENSE':
     Year: 2018-2020
     COPYRIGHT HOLDER: The PsychroLib Contributors

Please use YEAR: 2018-2020

   Found the following (possibly) invalid DOIs:
     DOI: doi.org/10.21105/joss.01137
       From: inst/CITATION
       Message: Invalid DOI

The DOI used in the doi markup is only 10.21105/joss.01137

Please fix and resubmit.

According CRAN policy, year field in MIT LICENSE should start as YEAR not Year. Also, DOI prefix should be removed.

This PR modified src/r/tools/deploy.R and src/r/inst/CITATION accordingly.

[dmey]

I'd prefer to keep master clean -- are you OK with waiting with the merge until CRAN accepts it? Just in case there are more changes to do...

[hongyuanjia]

Sure. Make sense. Will try second submission now.

[codecov-io]

Codecov Report

Merging #54 into master will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##           master      #54   +/-   ##
=======================================
  Coverage   88.42%   88.42%           
=======================================
  Files           3        3           
  Lines         380      380           
=======================================
  Hits          336      336           
  Misses         44       44

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 49a2baa...d511e68. Read the comment docs.

[hongyuanjia]

Finally get reply from CRAN maintainers (normally they will reply in a day):

Thanks,

Please omit the redundant "R" from the begin of the description.

"[...]Where copyrights are held by an entity other than the package 
authors, this should preferably be indicated via ‘cph’ roles in the 
‘Authors@R’ field, or using a ‘Copyright’ field (if necessary referring 
to an inst/COPYRIGHTS file). [...]"
e.g.: The PsychroLib Contributors, ASHRAE Handbook
Please explain in the submission comments what you did about this issue.

Please add references describing the methods in your package also to the 
description field of your DESCRIPTION file in the form
authors (year) <doi:...>
authors (year) <arXiv:...>
authors (year, ISBN:...)
or only if none those are available:  <https:...>
with no space after 'doi:', 'arXiv:', 'https:' and angle brackets for 
auto-linking.
(If you want to add a title as well please put it in quotes: "Title")

Please fix and resubmit.

For 1 and 3, it's easy to fix.

For the copyrights, it seems that there is no specific one acts as the cph role. I think it may make more clear to use an individual inst/COPYRIGHT file which contains contents like:

1. Hongyuan Jia and Jason Banfelder for R implementation.
2. D. Thevenard and D. Meyer for the core library implementations.
3. Equations and coefficients published ASHRAE Handbook — Fundamentals, Chapter 1 Copyright (c) 2017 ASHRAE Handbook — Fundamentals (https://www.ashrae.org)

What do you think @dmey @DJGosnell ?

[dmey]

this should preferably be indicated via ‘cph’ roles

this looks more like a preference than anything else -- In any case, let's try to keep this simple -- I would simply use what's already done at https://github.com/rstudio/rmarkdown/blob/master/DESCRIPTION#L36-L37. Specifically, I would change it to:

  person(family = "The PsychroLib Contributors", role = c("ctb", "cph"),
         comment = "Authors listed at https://github.com/psychrometrics/psychrolib/graphs/contributors"),

Then add you as maintainer

Maintainer: Hongyuan Jia <email>

[hongyuanjia]

That is a feasible way. My only concern is whether ASHARE should be put as an author.

[dmey]

That is a feasible way. My only concern is whether ASHARE should be put as an author.

OK, how about something like?

  person(family = "The PsychroLib Contributors", role = c("ctb", "cph"),
         comment = "Authors listed at https://github.com/psychrometrics/psychrolib/graphs/contributors"),
  person(family = "ASHRAE", role = c("cph")
         comment = "Copyright 2017 ASHRAE Handbook — Fundamentals (https://www.ashrae.org) for equations and coefficients published ASHRAE Handbook — Fundamentals Chapter 1."),

[hongyuanjia]

That is a feasible way. My only concern is whether ASHARE should be put as an author.

OK, how about something like?

  person(family = "The PsychroLib Contributors", role = c("ctb", "cph"),
         comment = "Authors listed at https://github.com/psychrometrics/psychrolib/graphs/contributors"),
  person(family = "ASHRAE", role = c("cph")
         comment = "Copyright 2017 ASHRAE Handbook — Fundamentals (https://www.ashrae.org) for equations and coefficients published ASHRAE Handbook — Fundamentals Chapter 1."),

Seems like this did not work. R CMD check directly fails. Probably we could keep the original style and also add ASHRAE?

E  checking DESCRIPTION meta-information ...
   Authors@R field gives no person with maintainer role, valid email
   address and non-empty name.

   See section 'The DESCRIPTION file' in the 'Writing R Extensions'
   manual.

[dmey]

Can you try simply by adding your email before role under "The PsychroLib Contributors"

[hongyuanjia]

No lock. It seems that the first "person" field should always have an non-empty name.

[dmey]

No lock. It seems that the first "person" field should always have an non-empty name.

How about you just add yourself as the first, rest should be fine as is?

[hongyuanjia]

How about you just add yourself as the first, rest should be fine as is?

Worked!

[hongyuanjia]

@dmey Latest reply from CRAN maintainer:

Thanks,


On 04.02.20 03:07, CRAN submission wrote:
> [This was generated from CRAN.R-project.org/submit.html]
> 
> The following package was uploaded to CRAN:
> ===========================================
> 
> Package Information:
> Package: psychrolib
> Version: 2.4.0
> Title: Psychrometric Properties of Moist and Dry Air
> Author(s): Hongyuan Jia [aut, cre]
>    (<https://orcid.org/0000-0002-0075-8183>), The PsychroLib
>    Contributors [ctb, cph] (Authors listed at
>    https://github.com/psychrometrics/psychrolib/graphs/contributors),

Authors and copyright holder should be part of the package.

If there are many authors and/or copyright holders, follow the advise 
and Writing R  Extensions and provide a file that lists all of them (one 
for authors and one for copyright hodlers), but they have to be part of 
the package as explained in the manual.

Please fix and resubmit.

According to Writing R Extensions:

The mandatory ‘Author’ field describes who wrote the package. It is a plain text field intended for human readers, but not for automatic processing (such as extracting the email addresses of all listed contributors: for that use ‘Authors@R’). Note that all significant contributors must be included: if you wrote an R wrapper for the work of others included in the src directory, you are not the sole (and maybe not even the main) author.

It seems that we have to list all authors like rmarkdown in a separate file and include it in the package under inst/ folder, instead of linking to a web file.

Will update and resubmit

[hongyuanjia]

@dmey It's on CRAN now! https://cran.r-project.org/web/packages/psychrolib/index.html

[dmey]

Great! But I am bit confused as the binaries are not available -- my understanding when we discussed the PR about having C++ source was that the package on CRAN would not require the user to compile psychrolib from source. Can you clarify this please?

[hongyuanjia]

It usually takes few days for CRAN to compile the package for Windows and MacOS. So just wait.

[hongyuanjia]

@dmey I believe this PR is ready to be merged. Right now Windows and macOS binaries have been built. You can check that by installing the package via install.packages("psychrolib").

[dmey]

Thanks for letting me know -- I will try to create the release and merge this in the next couple of days.