Change #1191712 had a related patch set uploaded (by Michael Große; author: Michael Große):

[mediawiki/extensions/GrowthExperiments@master] feat(ReviseTone): Show post-edit dialog if check was declined

https://gerrit.wikimedia.org/r/1191712

@KStoller-WMF @AAlhazwani-WMF Question about an edge-case:

Given that I clicked into an Edit Check from a Suggested Edit on the Homepage,
And that I revised the indicated paragraph somewhat
And that I try to save my changes
And that when my revised text is checked again and the model still finds a tone issue present
And that thus I'm shown again with a Tone Edit Check in the before-save stage
When I decline that Edit Check and then complete the survey,
Then ... ?

Should we still close the editor (but prompt the user whether they really want to close), or should the Edit Check just close as usual so that they can save their edit? I'd recommend the latter, but wanted to check in around that.

Thanks for thinking through these possible edge cases!
Given that they wanted to save changes and publish, presumably we should still allow the newcomer to save, so yes, the Edit Check should close as usual so that they can save their edit.

Side note: hopefully we track how often newcomers end up in loops like this. Ideally we avoid scenarios where this happens often, as I imagine it would feel frustrating and confusing for a new editor.

Change #1191712 merged by jenkins-bot:

[mediawiki/extensions/GrowthExperiments@master] feat(ReviseTone): Show post-edit dialog if check was declined

https://gerrit.wikimedia.org/r/1191712

Test Result

Status: ✅ PASS
Test Details: T401195#11279755

This task’s QA is included under T401195’s Revise Tone QA, which will cover the decline / skip behavior and flow.