ioos / bio_data_guide

Standardizing Marine Biological Data Working Group - An open community to facilitate the mobilization of biological data to OBIS.
https://ioos.github.io/bio_data_guide/
MIT License
45 stars 21 forks source link

Action to check "data help" tickets #169

Closed MathewBiddle closed 1 year ago

MathewBiddle commented 1 year ago

CF-discussions has established a way to use GHA to check issues that haven't been touched in x amount of time. We should do something similar for "data help" labeled issues in this repo (every 3 months?).

MathewBiddle commented 1 year ago

See https://github.com/cf-convention/discuss/blob/main/.github/workflows/stale-issue-reminder.yml

MathewBiddle commented 1 year ago

Ends up looking like this in the issue:

image

MathewBiddle commented 1 year ago

@albenson-usgs what would you like the message to be? Here's what I have so far:

This issue has had no activity in the last 90 days. This is a reminder to please comment on the status of the dataset and add any questions/blockers here. OBIS-USA node manager is also reminded to review @albenson-usgs @sformel-usgs

MathewBiddle commented 1 year ago

Also, do we want to add any tags to the issues? I would think this would be helpful in determining which to tackle first, but we don't want to ostracize folks... Tag could be "needs attention" or something.

Right now it just updates the issues with the label "data help" that are >90 days since the last comment with the text above.

MathewBiddle commented 1 year ago

It does this search

https://github.com/ioos/bio_data_guide/issues?q=is%3Aissue+is%3Aopen+label%3A%22data+help%22+updated%3A%3C2023-01-17

But, adjusting the updated date to be reflective of the days before today (2023-01-17 is 90 days before today).

FWIW, that search returns 9 issues that haven't been updated since 2023-01-17 (3 months ago). That seems like a lot. Should we think of expanding to 6 months (180 days)?

https://github.com/ioos/bio_data_guide/issues?q=is%3Aissue+is%3Aopen+label%3A%22data+help%22+updated%3A%3C2022-10-17

albenson-usgs commented 1 year ago

Nothing is ever simple 😂 It seems to me like this is going to need more thought/brainstorming before we implement. I do want to check in 3 months after the workshop because I feel like 6 months after the workshop will be too long. But the text we have right now probably won't make sense to people. Most people just put a placeholder issue in to let us know what dataset they are bringing to the workshop. Therefore it seems like we need workshop instructors to go in and comment on these tickets to start the conversation off. I don't think this can only be handled by me and Steve because it is too many tickets. I am also not sure what we should say to kick the conversation off. But I think whatever we say needs to wait until after the workshop.

MathewBiddle commented 1 year ago

At the top of the hour a test action will run on my fork for this process. I set it to debug so it doesn't do anything but report out a log of what it would do. This probably wont be informative because I don't have any issues on my fork 👎, but I'm curious what the log will say nonetheless.

https://github.com/MathewBiddle/bio_data_guide/actions/workflows/dataset-reminder.yml

I guess I'm not sure what you're looking for in this. I was thinking this was just a ping to remind folks that the issue exists and, if they have time, move the needle forward or say that they don't have time. I'm not sure it needs to be a customized message for each issue. Just something that says this hasn't been touched in x days, where are we?

albenson-usgs commented 1 year ago

Understood - that's why I think we need more discussion/brainstorming. It wasn't until I saw the message and looked at the issues that it would be added to that I realized the message wouldn't make sense to the people who would be receiving the message. I think that message would be ok after an initial reply from someone in the workshop team but if we were to add that message today to those issues, people would be really confused because they haven't worked on them at all and no one has said anything to them about them. So commenting on them "This issue has had no activity in the last 90 days. This is a reminder to please comment on the status of the dataset" would be confusing for the people who created the ticket because if I was one of them I would be expecting someone managing this repo to comment first and say...something...I'm not sure what at this point. I think it depends on what happens during the workshop with these tickets.