Proposal: Auto-close should leave different message if "needs info" label present

[platinumazure]

Example issue: eslint/eslint#11188

When the auto-close plugin closes an issue, it usually leaves a message saying we couldn't muster enough support for the change (with slight variations if the issue is accepted or not accepted). I think if the issue has the "needs info" label, we should leave a different/shorter message saying the issue was closed because not enough information was provided.

[not-an-aardvark]

👍, but it's worth noting that we sometimes forget to remove the needs info label from an issue even after info has been provided. It might be good if the bot checked whether the author added any additional comments since the needs info label was added, when determining which message to use.

(It might also be worthwhile to remove the "needs info" label when that happens, but probably not in this plugin.)

[nzakas]

Seems worthwhile. It should be easy enough to check if the last comment was left by a bot when the "needs info" label is present.

…

On Tue, Jan 22, 2019 at 2:11 PM Teddy Katz ***@***.***> wrote: 👍, but it's worth noting that we sometimes forget to remove the needs info label from an issue even after info has been provided. It might be good if the bot checked whether the author added any additional comments since the needs info label was added, when determining which message to use. (It might also be worthwhile to remove the "needs info" label when that happens, but probably not in this plugin.) — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#102 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AACWktaSBQjOsbl_7sooq9h2oejjauJ3ks5vF4x-gaJpZM4aNTih> .

--

______________________________ Nicholas C. Zakas @SlickNet Author, Principles of Object-Oriented JavaScript <http://amzn.to/29Pmfrm> Author, Understanding ECMAScript 6 <http://amzn.to/29K1mIy>

[platinumazure]

@not-an-aardvark @nzakas

👍, but it's worth noting that we sometimes forget to remove the needs info label from an issue even after info has been provided. It might be good if the bot checked whether the author added any additional comments since the needs info label was added, when determining which message to use.

I acknowledge this as a potentially bad user experience, but I really think we just need to be better about reviewing issues with "needs info" and removing that label if everything has been provided. On the other hand-- if an author comments with a complaint about providing info, or if other users help fill in the gaps, then we could have false positives or negatives very easily when relying on issue comments. I think relying on the label is the most sensible approach here, personally.

[nzakas]

This seems like something we could solve with the bot too: if "needs info" was added, and the issue author leaves another comment, we can automatically remove the label.

…

On Thu, Jan 24, 2019 at 1:34 PM Kevin Partington ***@***.***> wrote: @not-an-aardvark <https://github.com/not-an-aardvark> @nzakas <https://github.com/nzakas> 👍, but it's worth noting that we sometimes forget to remove the needs info label from an issue even after info has been provided. It might be good if the bot checked whether the author added any additional comments since the needs info label was added, when determining which message to use. I acknowledge this as a potentially bad user experience, but I really think we just need to be better about reviewing issues with "needs info" and removing that label if everything has been provided. On the other hand-- if an author comments with a complaint about providing info, or if other users help fill in the gaps, then we could have false positives or negatives very easily when relying on issue comments. I think relying on the label is the most sensible approach here, personally. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#102 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AACWkrg3CGGS62YKxZ1MrnCbaUl0VGeWks5vGibwgaJpZM4aNTih> .

--

______________________________ Nicholas C. Zakas @SlickNet Author, Principles of Object-Oriented JavaScript <http://amzn.to/29Pmfrm> Author, Understanding ECMAScript 6 <http://amzn.to/29K1mIy>

[platinumazure]

This seems like something we could solve with the bot too: if "needs info" was added, and the issue author leaves another comment, we can automatically remove the label.

I suppose, but it would also result in the label being removed if the issue author just leaves a complaint or otherwise non-helpful comment.

I'm no longer a team member, so take this with a grain of salt, but: I think it would be smarter to develop a search query that shows issues with "needs info" label where the last commenter is not the bot, and let a human take a look and see if the issue has the info it needs. Basically, right now it feels like the problem is the issue gets lost in the pile once info is added, and the only way to catch it as it happens is in notifications (not always feasible if notifications pile up).