Maybe we could borrow the UX from Minerva/MobileFrontend, where tapping thank gives you the opportunity cancel within a few seconds.

Does it do it? I think it’s bad UX design. Users may have slow computers, or they may not look back at the screen. Or they may not have enough time to react for physical reasons: by the time the screen reader announces the possibility to cancel, it’s too late. By the time someone who has difficulties using the mouse finds the cancel button, it’s gone. I’d like to keep the explicit confirmation, and probably bring that in Minerva/MobileFrontend as well. Making it more noticeable is probably a good idea, though.

There are two related tasks that go in a somewhat different direction, T282591: RFC: Proposed "unthank" feature to roll back "thanks" and T71636: Thanks: Implement an undo feature.

IMO it would be worth considering those tasks in the context of this work.

I think an "unthank" button or toggle would be more aligned with how users interact with "like" buttons on e.g. Twitter, Instagram, Facebook, etc.

In the case that a user presses "thank" and changes their mind, we could set some kind of grace period for X seconds before sending the email/web notification about a thanks, so the affected user would never see the notification. There would be some set of users where they do receive a "Thanks" notification/email based on an unintented action, or one that was later undone, but that doesn't seem like a huge problem?

On the "Current Design Explorations":

As touched on in the quoted T229168, "publicly viewable" isn't true at all, and likely deters people from using the feature. An accurate phrasing would be "The time and the usernames of both parties (the thanked and the thanking) will be recorded in a publicly viewable log, but not for which edit or on which page." That's pretty wordy, so a link to a page clarifying these matters would be helpful.

What will checking "Don't show again" do? Will it make you skip any and all confirmation? (I hope not, unless "unthank" is provided like discussed above.) Or will it show a more concise confirmation message? It's quite unclear.

My assumption is that improving the language and confirmation UX would be less effort than an undo/unthank option, but I purposefully made this task tile and acceptance criteria broad so we could explore multiple options.

In T315859#8548472, @Nardog wrote:

On the "Current Design Explorations":

As touched on in the quoted T229168, "publicly viewable" isn't true at all, and likely deters people from using the feature. An accurate phrasing would be "The time and the usernames of both parties (the thanked and the thanking) will be recorded in a publicly viewable log, but not for which edit or on which page." That's pretty wordy, so a link to a page clarifying these matters would be helpful.

Thanks for the feedback! I agree, we should brainstorm more accurate language. The copy is still a draft and we'll work on finding the right language that is accurate without being overly complex and confusing.

What will checking "Don't show again" do? Will it make you skip any and all confirmation? (I hope not, unless "unthank" is provided like discussed above.) Or will it show a more concise confirmation message? It's quite unclear.

The screenshot and Figma file have been updated. We likely either need the confirmation step to always appear or provide an undo option.

Adding the Figma file link here and will also update the task description.

For now we have two options, one with a dialog component and the other one with a pop up component. Prototype here.

Also some copy alternatives were explored which are all presented on the file.