Concluding discussions #313
Replies: 11 comments 4 replies
-
Repeating here a comment I made in another question:
|
Beta Was this translation helpful? Give feedback.
-
Thanks, @DocOtak. If you "mark as answer" does it close the question? I assume that anyone who opens a discussion can change its category subsequently (e.g. comment to question). |
Beta Was this translation helpful? Give feedback.
-
Hmm -- for #324 -- I only see "close" as an option, no "mark as answer" -- and the catagory is already "Q&A about using CF". |
Beta Was this translation helpful? Give feedback.
-
Ahh -- so that's a global (to the Catagory) setting, not a per-topic setting. So I'll let one of you change it if you see fit. |
Beta Was this translation helpful? Give feedback.
-
There are four types of Discussion format offered by GitHub: open-ended discussions, announcements, questions/answers, and polls. I created our three categories (announcements, comments and Q&A) all with the open-ended discussion format. This is the natural choice for comments. The sole difference between announcement format and open-ended discussion format is that discussions with the announcement format can be started only by a maintainer, not by anyone, and I didn't think we needed that restriction on our announcements. I didn't create a polls category because we've always decided things by consensus, not voting. We could introduce it of course if it were needed. I guess that "Mark as answered" is supported by the questions/answers format. I didn't choose that format for our Q&A because it has the upvoting facility, like you see e.g. in stackexchange. I don't think our Q&A are like that. Generally there are few answers, and often a discussion is needed to arrive at the right answer. Maybe we should use the questions/answers format. What do you think? I tend to think that it's easier to keep the same format, and introduce a label instead to "Mark as answered". We're accustomed to using labels effectively, I would say. |
Beta Was this translation helpful? Give feedback.
-
I have just been looking through the list of discussions and there are several that resulted in an issue that has been implemented, or otherwise are of little interest to keep open. I suggest they should be closed to avoid getting an ever expanding list of discussion that eventually become unmanageable. I suggest the following can be closed (no particular order): https://github.com/orgs/cf-convention/discussions/300 This is a suggestion and others might want to add to this list, or remove from it. |
Beta Was this translation helpful? Give feedback.
-
I'm not sure what the status is of the pull request mentioned in this comment, but until that is acted on I think we should probably keep. comment 322 open for now. I would not object to closing the others. |
Beta Was this translation helpful? Give feedback.
-
I'm not sure what the status is of the pull request mentioned in this
comment <#322>,
Looks like it’s been merged.
|
Beta Was this translation helpful? Give feedback.
-
In Lars's list, 362 and 339 are Q&A-type. I think there's a value in keeping Q&As open so that they will be noticed by someone looking for answers. They could be linked to answers in the FAQ as well. The others are of comment/idea-type. I agree they could be closed. The most recent update to any of those was on 23rd August, 3.5 months ago. Could we adopt a guideline of closing any discussion if it has been quiet for more than three months? Of course, if it has been clearly concluded, it could be closed earlier - or is there any value in keeping it open for a while after being concluded? |
Beta Was this translation helpful? Give feedback.
-
Does closing them make them harder to search / find? If so then yes, let's keep them open, but that's too bad. "asked and answered" would be a good category to have and be easy to search.
I don't think that's a good idea, ideas can sleep longer than that and still be useful -- a year, if at all? I don't know how hard it is to set up, but some projects have a bot that sends out an email to issue particpants (hopefully dicussions as well) saying that an issue has been dormant for some time, and that it will be closed if no one revives it -- that would be4 a good idea -- anyone have the github-fu to make that happen? [in the meantime, I suppose we could do it by hand] |
Beta Was this translation helpful? Give feedback.
-
Here's the status, as I see it: I took it upon myself to close a few that seemed clearly ready to be closed -- they can always be re-opened if I made a mistake: https://github.com/orgs/cf-convention/discussions/300: concluded -- I've closed it. https://github.com/orgs/cf-convention/discussions/320: "Broken web links" -- This looks not totally resolved, but should probably be turned into an issue for any remaining ones. -- I posted to that discussion asking for resolution https://github.com/orgs/cf-convention/discussions/322: this has been resolved -- more to do to improve it (links to CF projects) -- but I don't think we need it anymore. -- I just closed it, but someone can re-open if they think that it needs more attention. https://github.com/orgs/cf-convention/discussions/310: I think this is done, but leaving it to the standard_name gurus to close it. https://github.com/orgs/cf-convention/discussions/342: this is about the 2024 CF workshop -- closed it. Questions:https://github.com/orgs/cf-convention/discussions/339: question: keep open until / if a FAQ entry is added https://github.com/orgs/cf-convention/discussions/362: question: keep open until / if a FAQ entry is added |
Beta Was this translation helpful? Give feedback.
-
Topic for discussion
I wonder how our GitHub discussions should be dealt with when they stop. Some of them will be aware of coming to a conclusion, and others will just go quiet (like issues), I suspect. Here are a couple of questions:
Should we close them when they've stopped (for whatever reason) or leave them open? I think the only difference this makes is whether they appear in the default list. It's sorted by latest activity, so the inactive ones will gradually go down the list. With issues, there's a reason to close them, namely that the list of open ones compose our work agenda, but I don't think that argument applies as much to discussions.
Should we label them, as we do with issues, e.g. as concluded or inconclusive, if they are conscious of having reached an end of some sort? We could also label them automatically as dormant if they have been quiet for some time without being explicitly concluded.
Best wishes
Jonathan
Beta Was this translation helpful? Give feedback.
All reactions