Hacker News Reader

> It's a human problem, and like most problems of this genre there is no "best" answer

For this specific problem (Google Meet UX/UI) sure anyone agree with: don't put a destructive action that requires no confirmation (leaving the meeting) next to a common action (mute/unmute yourself). If designers don't get that right, sorry but they are not competent designers. It's like a programmer that, in order to "balance enormous numbers of competing constraints and concerns" decides to not escape user-provided HTML in the frontend. Well, that programmer is not a competent one.

Another classic case of this is your car's "key fob" buttons. You usually have three functions on three similarly sized buttons right next to each other: lock doors, unlock doors, and... SOUND THE DEAFENING ALARM. Really? Can anyone spot the one that doesn't belong next to the others?

On my key fob the alarm button is small, red, and inset into the side, and it requires a lot more force to push. I've never pushed it accidentally. On the face are Lock/Open trunk/Unlock, but it's not obvious that you have to hold the trunk button for two seconds for it to work (this is probably because you have to manually re-latch the trunk if you unlock it).

> next to a common action (mute/unmute yourself)

I agree with this point, but if the designer had a phone call in mind (with a pre-pandemic mindset), then the design feels more reasonable.

In a one-on-one call, 'mute' is much closer in intent to 'hang up' - i.e. I am not currently participating in this call.

I agree with you about the tradeoffs and the importance of looking at the problem from all angles, but I would argues that this is clearly the case wher aesthetics won over usability, which should _never_ happen. It should be easy to visually separate the "hang up" button from other, much less destructive and more commonly uses buttons (especially "mute"), for example with some small amount of separating space. The only reason this is not done is that it "wouldn't look balanced" and designers would be unhappy.

In my experience designers are the worst people to hire for UX because they often sacrifice usability for aesthetics. Programmers fare a bit better (usable UI usually doesn't conflict with the code quality), still not perfect though. Casual users are probably best, with some education in usability of course.

Yes! A comment from a few month ago resonated with me: devs are to blame for mediocre design, designers for terrible design.

https://news.ycombinator.com/item?id=24760194

Completely agree. And the thing I love about your comment is that even though you've found solid argument for why the OP's criticism was right, you articulate the tradeoff and explain why you think it was the wrong one. This is a dramatically more effective and compelling approach IMO.

If you happen to do design work and are looking for a job, I'd love to chat and see if there are any possibilities for collaboration. If you're interested, feel free to drop me a line at my username at google's mail service.

Zoom solves this problem by.... adding a confirm popup button. As far as I can tell they're the only meeting app that does that, I find it almost annoying most of the time but very useful when I hit it by accident.

I've often wondered why this kind of confirmation couldn't be conditional on the confidence that the click / touch was intentional?

As an example, my favourite pet peeve was using Visual Studio with old-school, upfront locking source control (like TFS), and then accidentally drag and dropping a file or folder due to lag in remote desktop, or a failing mouse which sent two click events in 1 ms or something. VS duly pre-emptively locks the 10k files in the folder you just dragged, and begins a 5 minute operation you'll have to somehow undo later, even though it should be fairly obvious from the click events that it was non-intentional.

Going back to the meeting example, surely solid, accurate taps in the center of the hang-up icon could be taken as intentional, but a kind of glancing, less accurate one needing confirmation?

That type of approach is the way forward in UI. Maybe if we call it something silly like MLUX it will catch on sooner.

Another instance of the same principle is that if an unexpected button/element appears and I click it in <30ms or whatever the fastest possible read+react time is, the click isn’t intentional and should be ignored or confirmed. This should scale over time based on user familiarity and speed.

This culminates at:

> as the technology became more sophisticated the controls were made touch-sensitive - you merely had to brush the panels with your fingers; now all you had to do was wave your hand in the general direction of the components and hope. [HHGTG]

Changing the way buttons behave that is completely opaque to the end user is a good way backwards.

> if an unexpected button/element appears and I click it in <30ms

Easily solvable by locking the buttons for a few seconds (or less).

Disadvantage there is advanced users know where the button will land and have to wait.

That popup annoys me, because I never hit the button to leave a call by accident. It used to annoy me even more because there was a time last year when it was impossible to leave a call by keyboard only on Windows, because all methods of normally closing the window would trigger this popup, but you couldn’t submit that popup by keyboard. Now at least I can Alt+Q Enter or something like that.

(I confess that the first time I encountered this I opened Task Manager and found and killed a Zoom.exe process all by keyboard, just on principle. That was when I discovered that Zoom has two processes so that the main one can restart the call if the call process crashes!)

But you know one potential factor for my never hitting it by accident? They use text labels (“Leave Call” / “End Call”) rather than an icon. Much easier to get right.

It feels like one of those things where the context of the meeting/call should be taken in: was the scheduled time of this call for an hour and you've left 10 minutes in, maybe a confirmation dialog is more useful than when you've decided to leave within a few minutes of the finish time. Doesn't cover all the edge cases etc., but just observing that "the right UX" is super contextual and not a one size fits all thing.

Whether taking that context into account is a good idea is very debateable.

The more contextually 'intelligent' a system is, the harder it is for the user to model it, and ease of user modelling is often more important than reducing the number of interactions need to complete a task.

In this particular case, it wouldn't be possible to know if your keyboard sequence that quits a call would feature an unnecessary enter at the end or not.

Doesn't fix the problem of accidentally turning your video on when you meant to shut the call off. I've taken plenty of calls in places where it'd be more professional to keep the video off (especially when outside home)

> The thing the author doesn't seem to realize / acknowledge is that UI/UX design is about balancing enormous numbers of competing constraints and concerns. It's a human problem, and like most problems of this genre there is no "best" answer, only different sets of weights for different, often competing, concerns.

To underline this point, I think in general design suffers from a lot of bikeshedding at tech companies. There are likely many designs tucked away in discarded files that addressed this specific pain point, too, but were discarded at the request of some PM, or manager, or director. Then there's the process of user testing and experiment design that is used to validate UI/UX of products like this. This design may have actually tested well even if it wasn't the team's favorite... I've worked places where that data is used to override a designer's opinion.

Now apart from all that of course, Occam's razor probably applies as well I guess... perhaps this bit of UI was just poorly designed. But I see a lot of chatter on HN regularly about design being superfluous, subversive, unintuitive, "bad" when really I think many (most?) designers are unhappy with the designs that ship out as well.

There’s also the issue of trying to balance what you know to be good vs the desires of the person actually paying you. Fighting this directly doesn’t get you very far (you’ll most likely be replaced by someone more pliable) and some people cannot be convinced of some things.

Everyone and their mother has an opinion on what they think is good design. I am thankful that as a developer I am not having to fight over pixel pushing.

I think the trade off here is that if you don’t put the end meeting button there, users will get frustrated because they can’t find the end meeting button.