This patch provides the response and reporting guides for Code of Conduct reports. It also removes the word draft from all the documents,
adds information about the CoC committee, and a place to put transparency reports. A post will also be made on Discourse to provide more details about the history of the code of conduct, this patch, and next steps. Please see that post for full details. I will put a link below once I have it.
Details
Diff Detail
- Repository
- rG LLVM Github Monorepo
Event Timeline
llvm/docs/ResponseGuide.rst | ||
---|---|---|
93 | Is the double empty line intentional? | |
128 | Does that include forbidding someone to send a patch to the project for example? What if this person published a patch, can someone else still cherry-pick it in the repo with the author attribution? (I would think that a very extreme situation would be needed to trigger such a response anyway) | |
130 | Is the double empty line intentional? | |
178 | Formatting nit: this may not be very important, but this file does not have an empty line after the section titles. |
llvm/docs/ResponseGuide.rst | ||
---|---|---|
106 | "CoCc" is not used anywhere else in this patch. | |
108 | Should this be business days with consideration for the locality, etc. of the reportee? | |
127 | I'm not sure what style guide supports it, but other instances in this patch uses "ie.". It seems to me that most of the uses introduce examples though (so "e.g." may be more appropriate). |
llvm/docs/ResponseGuide.rst | ||
---|---|---|
108 | I find "business days" to be quite un-inclusive actually, in particular for OSS projects. Not everyone is paid to work on the project and if someone reports something on a Monday, they may only be able to dedicate significant more time on their week end for example. |
Thank you for working on this, I'm really glad to see the CoC coming out of DRAFT status and being finalized! I think these documents are going in a great direction, just found some minor points/questions.
llvm/docs/CodeOfConduct.rst | ||
---|---|---|
98 | Please wrap to the usual 80-col limits and convert smart quotes like ’ into dumb quotes ' (here and elsewhere in your changes). Do we want to make conduct@llvm.org into links in this document so people can just click it if needed? (I don't believe that will add any additional spam burden on the account -- listing the address is sufficient to trigger spam.) please see our Reporting Guide -> please see the Reporting Guide | |
114 | Rather than having people go to a second source, it would be nice to list the members here directly. It's a bit of an extra maintenance burden, but who a CoC report is seen by is crucial information when filing a report, so it seems more user-friendly to not hide that behind another link. | |
119 | Do we want to add any extra information about what's expected to be in the transparency report (or more importantly, what's expected to NOT make it into a transparency report) as not all readers may be familiar with this? I see later that we do have some information on what's in the transparency reports elsewhere, so maybe we want to link to there from here? | |
llvm/docs/ReportingGuide.rst | ||
66–67 | Similar changes are needed for other section titles (so long as it's consistent within the document, what style you go with doesn't matter much to me). | |
66–67 | Same question here about whether we want to make a link for the email address. | |
67–68 | How does this change confidentiality of the report? e.g., should the reporter assume that in addition to the CoC team, all staff members have access to the report, or only the staff member contacted, something else? | |
70 | I think this policy is reasonable, but it does leave some daylight for reports to be accidentally dropped if the meetup organizer doesn't pass the information along to the conduct team. Would it make sense to modify this slightly to report the incident to the local meetup organizer AND CoC team at the same time, with the understanding that the local organizer should take immediate action and report back to the CoC team. This way, the CoC team is still alerted to the potential issue and it helps mitigate the risk that an event organizer doesn't report up (either accidentally or otherwise). | |
71–72 | Should we add a bullet point discussing anonymous reporting? | |
71–72 | ||
llvm/docs/ResponseGuide.rst | ||
20 | It's not entirely clear that "listed on the LLVM website" is quite accurate. For example, we have working group meetings listed on the website, as well as community office hours, and neither of those will have a CoC response team. We may want to reword it slightly to say that there MAY be a code of conduct response team, but if there's not one specific to the event, file the concerns the usual way. | |
41 | I <3 Oxford commas, no idea if others love them as much as I do. | |
60 | Do we have a retention policy we want to mention for how long we keep all this data? (I read this as "we retain it forever".) | |
108 | Given how many three-day weekends exist in different places, I agree that three days may be a bit tight and seven might be a bit more palatable. However, I'm also not opposed to three days because I understand the need for addressing the incident in a timely manner. But I'd avoid "business days" (that also gets into things like whether a national holiday is or isn't a business day, etc). | |
116 | ||
143 | Do we want to nail "reasonable time frame" down further? I'd have guessed we'd want to guarantee some resolution within some time period (two week goal, as with the original report?) |
I have addressed all the comments or responded why I did not make the change. I will be updating the patch next.
llvm/docs/CodeOfConduct.rst | ||
---|---|---|
98 | Those style quotes are required to make the links work correctly. Anything with @ is automatically turned into a email link. I made the our/the change. | |
119 | This section is meant to just list them, but I will add a link to the description. | |
llvm/docs/ReportingGuide.rst | ||
66–67 | Sphinx generates this into a link for the html pages automatically. | |
67–68 | I will modify this to make clear that the staff is there to help you find a Code of Conduct team member and not resolve the incident. | |
70 | Sure, I will add this. | |
71–72 | The process does require a working email so we can get more information from the reporter and get feedback on the resolution. It doesn't mean that someone can't create a new email and exclude personal information. However, I don't know if we should actively encourage anonymous reports either by explaining how to do it. | |
llvm/docs/ResponseGuide.rst | ||
20 | Well, this is talking about in person events. AFAIK no working groups are in person right now. But it will apply to other events such as workshops at conferences. This will be documented either on the website or in docs. However, all working groups should have a response/contact person(s) clearly listed on the website for multiple reasons (including CoC immediate response). I plan to contact all working groups to gather this information in the coming weeks as it is not listed currently. Office hours will also need to mention this somewhere and right now that is on the webpage, but it should also be on the calendar listing as well. | |
60 | In my opinion, it's going to depend on the violation and resolution. As the CoC committee will be changing over time, some data is important to keep longer than others. I think the CoC is going to have to figure this out over time. | |
93 | Nope. Will be fixed in updated version. | |
108 | I will change this to a week as I agree it would meet more schedules. | |
128 | Most likely yes. All of our methods to submit a patch to the project go through some sort of infrastructure - mailing list or Phab. So their access to those things would be removed through a ban. If someone gets their permission and wants to take over a patch as their own then they can do so. However, they should not act as a proxy for someone who is banned. This would now be their patch. This is a very specific situation and if someone is given a ban, the CoC committee would provide the actual details of the ban and have final say on this. | |
130 | No, will fix. | |
143 | Sure, I can change this. | |
178 | Ok, will be fixed in next revision. |
I'd like to get this finalized this week, so if past reviewers are happy, please accept.
Aside from the style nit with smart quotes, I think this looks great (feel free to fix those when landing instead of going for another round of review). Thank you @tonic for all the hard work that went into this!
llvm/docs/CodeOfConduct.rst | ||
---|---|---|
98 |
That's not the quotes I was asking to have changed, so I'm sorry for the confusion. e.g., This isn’t a public should be This isn't a public (We don't use smart quotes in other documents, so this is a change for consistency.)
Ah, great to know, thanks! | |
llvm/docs/ReportingGuide.rst | ||
71–72 |
That's a fair point; we can circle back later if it turns out we find a need to add more details here. | |
llvm/docs/ResponseGuide.rst | ||
20 | Excellent, thank you! | |
60 |
Okay, that sounds like a good reason to not nail it down in the docs, thanks! | |
202 | This is another instance of smart quotes: “take a week off” -> "take a week off" |
llvm/docs/CodeOfConduct.rst | ||
---|---|---|
98 | Ah, got it. I will fix those. |
Got halfway through so far. Some editorial comments.
llvm/docs/CodeOfConduct.rst | ||
---|---|---|
95 | Minor: Spelling correction. | |
llvm/docs/ReportingGuide.rst | ||
-29 | Minor: Fix mid-sentence double space. | |
-34–1 | Fix possible typo that causes an incomplete sentence. | |
0–1 | Minor: Formatting fix. | |
0–1 | Minor: Many style guides would not capitalize "for" in title case. | |
2–3 | Fix formatting. | |
llvm/docs/ResponseGuide.rst | ||
32 | Unaddressed instance of Aaron's comment about applying title case consistently. |
Some more editorial comments and a comment regarding some text that does not seem to convey whatever meaning was intended well.
llvm/docs/ResponseGuide.rst | ||
---|---|---|
41 | I guess we aren't using "Oxford commas"? | |
44 | Prior comment missed? | |
66–69 | Prior comment missed? I still don't know what this is trying to say. Organizers should use discretion in determining whether a talk should be ended early if Code of Conduct violations other than (any single instance) of harassment or violent threats has occurred? What does the "or they are infrequent" refer to? Is the intent that the sentence about harassment of violent treats applies only to repeated or ongoing cases? | |
75 | Minor: Title case. | |
164–165 | Prior comment missed? "CoCc" is not used anywhere else in this patch. | |
189 | Inadvertent new bullet point? | |
261 | Minor: Use lowercase "the", which would be consistent with other similar uses in this patch. | |
290 | Inadvertent new bullet point? |
So sorry, I think my ResponseGuide changes got reverted. I am going through these again to make sure they actually got completed.
llvm/docs/ResponseGuide.rst | ||
---|---|---|
66–69 | Sorry, some changes got reverted accidentally. |
@hubert.reinterpretcast - Can you accept if you believe I have made all your requested changes?
There were some minor issues still left. LGTM whether they get fixed or not though.
llvm/docs/CodeOfConduct.rst | ||
---|---|---|
94–150 | Minor: "commitee" should be "committee" instead. | |
llvm/docs/ReportingGuide.rst | ||
-34–1 | Minor: Should be "in-person" in place of "in person" here. | |
llvm/docs/ResponseGuide.rst | ||
66–69 | Not a problem; thanks for fixing. |
Minor: Spelling correction.