I've seen this pattern in the wild already. IIRC my bank (Barclays UK) has a section in their marketing email preferences where you need to check the box IF you don't want to receive any marketing^Hspam email. People notoriously hate to read, and even grandmas have learned the lesson never to check any box at the end of any form, and this is how they still get you.

It's atrocious and there should be legal consequences to whoever adopts or implements dark patterns like these.

On the topic of hating to read, it reminds me when I bought my TV, the online store would only sell it on credit, with 0% interest. The paperwork came in the mail with a free credit card too. The sign up form (from some bank, not the store itself) had 2 signature fields, but the 2nd field was to sign up for insurance tied to the card. I looked it up online and there were so many mad reviews of this bank from people who signed the 2nd field and claimed they got scammed...

This is different. The text say yes, but the x says no.

I've seen worse. Two marketing options buried at the end of a form. One option you have to tick to opt-out, the other you have to untick to opt-out.

Both written in a confusing way, and you get your dara shard with third parties and marketing of some kind if you get either wrong.

It is against the GDPR, but this of course lacks teeth.

I always found that frustrating. In some Windows apps, there is a toggle switch with words like: 'Disable {this}' and the toggle is turned off by default. So which is it? Enable or disable? One trick I learned is that if there's a double negative in a user interface, then that means a positive. My computer science course that involved learning about logic gates helps me out in these situations.

Often found in Group Policy settings, as explained by Raymond Chen: https://devblogs.microsoft.com/oldnewthing/20110606-00/?p=10...

Yeah, toggle switches for some reason are always infuriating. I particularly hate the ones with text inside. It appears ticked, but says 'Disable'...wait is it disabled or telling me to click to disable. I much prefer radio buttons.

There is a similar pattern in programming where some folks insist on Boolean variables.

Guessing this is a troll comment, true or false?

Maybe if it was a button, but not if it's a switch.

That one is clear and unambiguous though, there is just an extra layer. Disabling it is disabled.

It's worse when they change the text when you toggle it, so either "Disable x" is disabled or "Enable x" is enabled, both of which mean the same thing.

I've been enjoying watching Neil's hacks a lot. Incredibly creative ways of messing with Mac display and libraries. I encourage folks to scroll through his Twitter gifs for some great hacks.

Edit: I don't even know if they're all real or are just effects but either way they are all incredibly creative.

I've seen that for small icons. The "off" state is the icon for the thing, and the "on" state is something like two vertical bars, regardless of what was turned on.

He also wrote:

"Lol imagine if files and folders got heavier the bigger they get"

Someone did that in the early Mac days, as the "Sonic Finder". Everything the use did made a sound. Moving files produced a "thump" sound, with a deeper "thump" for larger objects.

Few people left that feature turned on.

I wonder if this could reasonably be applied to the rm command - something like "wait for log(N) seconds before actually deleting N files."

For people looking for more, this sub-reddit(1) is just filled with such bad/evil UI components, though most of them are more bad than evil.

https://www.reddit.com/r/badUIbattles/

I think what's about as evil and actually widespread, is the UI component which functions as a check box, but consists of a circle in an oval track, that goes from left to right and vice versa when you click it.

One can assume that left is off/no and right is on/yes, but if someone should want to reverse the meaning, how can you prove that's invalid? Especially if the text is ambiguous in the right way.

You can't assume the default position is off or on, because that varies.

And changing the color depending on the position creates more uncertainty - the user can tell that it means something, but what?

I trained my junk mail filter to automatically mark as junk anything with "unsubscribe" at the bottom of an email. As long as this is used to make subscribing to digital garbage only, I'm fine with it.

You see this a lot in the RT ticket system[1]. For example, when adding a reply to a ticket, emails are sent to a set of people by default. The addreses are listed, with checkboxes, and a parenthetical Check boxes to disable notifications to the listed recipients

[1] https://bestpractical.com/request-tracker

Will not work in Sweden as "x" means the same as "v" on the paper

And some people use a checkmark for "wrong" for some reason

If it were truly inverse, the text should also be inverted, for example "I Opt Out Of Emails." And it should start Xed, so that enabling it is like closing the Off. Then I would be fine with it.

Wow this thing is such an abomination that about half the commentators in the original thread literally are taking home the opposite message (ie ‘x’ to confirm vs checkmark).

I don't know, if it was a paper checkbox I would check it with a cross (x). Like on an election ballot. Is the X just a stylistic change or is it supposed to mean something different than an unstyled checkbox?

The red X means you _don't_ want what you ticked.

Seen this pattern in tracking consent popups - active state is red, making you think it’s disabled.

I think this would now be illegal in California...but enforcement would be tough.

Segmentation for professional manipulators who like their job with a challenge...