Community Sustainer (Moderator) Guide
An onboarding guide for community.rstudio.com folks that have signed on to become community moderators.
First and foremost we’d like community.rstudio.com to set an example for being a welcoming and kind place for discussions around R, RStudio’s toolchain, and data science.
We’d like to emphasize setting a good tone, being welcoming to people new to R and applied statistics.
We’d like to keep snark to minimum (certainly in public forums).
Documents to be familiar with
There are a few pages I’d love for you to be familiar with and comment on.
Understandably, most users are unlikely to ever read our guidelines, FAQs, or Code of Conduct.
Get to know our existing FAQs. The hope is for these to be useful links to supply when people have common issues.
Guidelines motivating our community: https://community.rstudio.com/faq#civilized
Frequently asked questions; https://community.rstudio.com/faq
Code of conduct (coming soon), likely to be added to guidelines. (Likely based on https://wiki.r-consortium.org/view/R_Consortium_and_the_R_Community_Code_of_Conduct and https://github.com/todogroup/opencodeofconduct/blob/gh-pages/README.md)
Big takeaway, please avoid posting personally identifiable information (full names, emails, etc.) in topics and posts. Why? unlike in profile summaries, this can be hard to find and remove later.
Please offer feedback
Many of us use community in very particular ways, or get used to certain behaviors, so feel free to bring up potential issues even if you don't have a suggestion to fix things, or are clear about what the problem is.
There's always room to improve. Where you think it is needed, please give feedback on existing guides and FAQs. (As an example.) But also, please don’t be offended if others disagree or it takes a while for anything to actually change, it happens. Building consensus takes time.
- Public Discussions - For most of these discussions go to #meta.
- Private Discussions - If there’s some reason keep the topic private (discussions about specific users, for example) then turn to private messages or the #lounge.
- Flag discussions can be a good place to offer feedback as well (some more details below).
What kind of feedback?
- Is someone exhibiting a pattern of behavior that you think is problematic? Perhaps we should discuss and create have a policy to curb it.
- Perhaps an existing guideline is too weak or too strong. Perhaps a guideline is too complex.
- Have you repeated some advice a few times? - let’s chat about creating a guide (or package or video) future people can point to.
Community "Help Assets" vs Outside Content - if a resource should be created with advice specific to this forum, I feel we should host it via a #meta wiki. Otherwise, we have a general preference to point to as many outside resources -- preferably created by community members -- as possible.
Some suggestions for dealing with discussion issues
Recommendations for handling issues and problematic behavior on community.rstudio.com.
Reply with suggestions - Initially, please reply publicly with suggestions (e.g. "please ask this as a reprex").
- This (hopefully) establishes good norms.
- Canned responses as part of your reply can be handy.
- There are a number of canned replies with links to FAQs already.
- Flag - If you’d like someone to take a look, flag the topic and at least Curtis will check it out within a day.
- Send a private message - Sometimes, for the sake of not distracting attention from the thread’s topic or creating an embarrassing situation, a kind private message can be effective. This may be tedious and may open you up to a difficult conversation. Feel free to suggest Curtis handles this by flagging a topic.
Hide/Unlist a topic or reply - If you think it’s really important, you are able to hide a post.
- Do this by flagging it, and then selecting “agree and hide post” (note that the user can see who has hidden their post).
Features to be familiar with
Community.rstudio.com is run on discourse.org
- We use discourse to run our forum. Discourse is an open source forum software, we use the hosted service.
- Discourse was founded by one of the founders of stack overflow, but where SO is focused on the FAQ, answers and incentives around answering questions -- discourse is focused on 1) more free-flowy discussions 2) more focused on users (who have an identity/profile and point of contact).
- Many discourse threads may turn into solid FAQs that compete with SO answers (or r-help, quora, etc), but the site isn’t really optimized for that as well as SO is. But we should be aware that some threads are ranked high in important search results.
#Meta and #Lounge discussions
We'd like to continuously improve our FAQs, our guidelines, our canned responses, and other documents we push to users frequently.
If you ever have an idea, or some doc strikes you the wrong way, or maybe could be split, let’s talk about it.
Flagging a topic or reply brings it to the attention of other sustainers and the site’s admin staff.
Flagging is an efficient way for you to highlight an issue, while not obligating you to spend a lot of time resolving it.
For example, Does someone repeatedly ignore pointers to our homework policy?, or requests for a reprex?, or misses the tone we’re trying to set? Does a post suggest an issue with some guideline, or the need for a better FAQ or canned response?
Some flag facts:
- Low flag counts don’t hurt the user.
- Flagging is (mostly) anonymous. Users can see that they have been flagged, but not by whom (unless they are a sustainer).
- Discuss. Flagging opens a discussion among sustainers. Feel free to offer your two cents. Maybe we need a better policy or resource? Maybe we need to ban this person?
“Red Alerts” - Three types of alerts
There are three "red alerts"
- Please leave for Curtis to clear.
- Flagging creates a group private message with all moderators. Think of this as a Moderator Inbox. You’ll see “ Red Alert “ flags.
- Feel free to comment or ignore as you see fit. But for the time being, please leave clearing flags to Curtis
- Feel free to clear
- This is a spam filter for comment and new topics. https://akismet.com/
- Feel free to clear
- A spam filter run by discourse, based on data from all the sites they host. (e.g. this user email/IP was posting spam over here, so it’s probably spam on your site too)
How to Flag
- Below the reply to be flagged, click the “...” ellipses to reveal the flag icon.
- Click flag
- Select “Something Else” and add a note with some context
Something Else Examples:
- Does this violate our HW policy?
- Has been asked to give reprex and format his posts, but ignore.
- This is a little too snarky for my taste.
Discourse has a canned response tool. This tool is a list of canned-responses from which you may copy-pasta text into your message. It can save a lot of time. You can also contribute to the list.
When you open a Reply Message, or Private Message click the gear icon, then “Canned Responses”.
This opens a canned response menu. You can see other sustainer canned response and create your own. Feel free to use your own voice and cite your own preferred content.
Spread Love and Mana
Likes / Hearts
If you ever feel inclined, click the heart button to like others replies. This signal is used to indicate how helpful folks are on the site. This signal is particularly strong when it comes from you.
We are worried about sustainer burnout. We’d hate for you to spend time tediously dealing with problematic people or issues, or feeling obligated to do anything you weren’t otherwise highly inclined to do anyway.
But also, ignoring issues can lead to problematic behaviors becoming norms. Or lead to things not improving here.
Hopefully these features are easy to use, and help fulfill our goal of being a welcoming, kind, and useful place.
Discourse moderation guide: https://meta.discourse.org/t/discourse-moderation-guide/63116
Includes discussions of discourse features, and ideas for additional interventions for problematic behavior.
Art Of Community by Jono Bacon. A nice review of working with and running online communities. http://www.artofcommunityonline.org/
A Theory of Moderation by Jeff Atwood on May 18, 2009 https://stackoverflow.blog/2009/05/18/a-theory-of-moderation/
Note on tone - sometimes jokey or clever isn’t wise.