dev_guide Clarify section on Code of Conduct?

Our advice (and reviewer checklist) asks for the code of conduct to be linked and suggests using the usethis function, https://devguide.ropensci.org/collaboration.html#friendlyfiles. This means that most developers will use the default usethis code of conduct. Two questions:

Wouldn't it be better for us to tell authors to add a link to ropensci's code of conduct at https://ropensci.org/code-of-conduct/ ? If not, then I'm a bit confused about the scope of both codes. I know a 'contributor code of conduct' is a bit narrower in scope than the the ropensci code of conduct, but it seems the latter should cover the former as well, and seems kinda like I'm opting out if I'm declaring my package to be governed by a separate code of conduct (with no constraint on what I'm expected to actually list there if I do deviate from the usethis template).
The usethis function directs the user to paste this block into the README:

[Contributor Code of Conduct](CODE_OF_CONDUCT.md).

I always paste this exactly, and then almost without fail I get the package rejected from CRAN for having a URL that doesn't resolve. (This is fixed in the dev version but for now most users probably still have the CRAN version of usethis, https://github.com/r-lib/usethis/pull/917)

Jan 24 '20 17:01 cboettig

Thoughts from our code of conduct committee team @stefaniebutland @sckott @karawoo appreciated here too!

Jan 24 '20 17:01 cboettig

hmm, the one on our website is more comprehensive. we do want folks to have a code of conduct file in their repo for sure though. maybe we could suggest a COC file that links out to the COC on our website?

yeah, the coc url instead of path link should be made clear - i don't use usethis, but good to see they're updating it.

Jan 24 '20 17:01 sckott

cf also https://discuss.ropensci.org/t/using-ropensci-code-of-conduct-on-github-project/1898

Jan 30 '20 07:01 maelle

cc @Bisaloo

Jan 30 '20 07:01 maelle

@stefaniebutland @karawoo would it make sense if the dev guide stated the COC after transfer should be a link to rOpenSci COC?

Jan 30 '20 07:01 maelle

In the absence of response above, I'm removing this item from next release. Thanks for your understanding.

Feb 06 '20 14:02 maelle

I've opened an issue for the Code of Conduct Committee to make a decision and specific recommendation. This is not likely to happen before May (sorry).

Mar 04 '20 17:03 stefaniebutland

Sorry it took me a while to wrap my head around this. I agree with the current suggestion to link to rOpenSci's Code of Conduct. Since Scott also agreed I don't think this requires further discussion with Kara since it doesn't influence the Code of Conduct itself.

Mar 26 '20 02:03 stefaniebutland

Thanks @stefaniebutland and @sckott.

@sckott (and anyone who wants to chime in) a few questions

so in practice after approval of the package should CODE_OF_CONDUCT.md content be

Refer to [rOpenSci Code of Conduct](https://ropensci.org/code-of-conduct/)

?

And what about the COC before review? Do we require one? Here's the current content on COC in the dev guide https://devdevguide.netlify.com/collaboration.html#code-of-conduct-1
what about existing code of conducts in packages that have been onboarded a while ago? Do we want to do a PRs campaign to update them? In practice we could delete the existing code of conducts actually, if we use the approach suggested by @Bisaloo (having a repo ropensci/.github with the COC). We could start the .github repo with just the COC and see about contributing guide / issue template / PR template another time.

Mar 26 '20 07:03 maelle

https://github.com/search?l=&q=user%3Aropensci+filename%3ACODE_OF_CONDUCT.md&type=Code

https://github.com/search?l=&q=user%3Aropensci+filename%3ACONDUCT.md&type=Code

:grimacing:

Apr 02 '20 14:04 maelle

what's in your PR looks good

Apr 06 '20 15:04 sckott

A good first step would be to update the approval template so that newly approved packages follow the latest guidelines :wink:

Aug 31 '20 20:08 Bisaloo

That's a good suggestion but I'm a bit undecided: if new packages were built using usethis, or GitHub docs, they'll most likely already have the right filename. And the approval list is long. So probably better as an editor check :thinking:

Sep 07 '20 07:09 maelle

wait I got confused, I need to update the approval template indeed :woman_facepalming:

Sep 11 '20 11:09 maelle

The approval template had been updated.

Now what's needed is making sure old packages follow the new guidelines.

Jan 12 '21 12:01 maelle