Hacker News Reader

I factory reset my phone yesterday and during the setup process Google asked me to enable location & web tracking. There was a thick, blue button that said "Turn On" at the bottom. But until I scrolled the arbitrarily extended page all the way to the bottom, the option to keep it off (with no button background, just text) didn't appear.

That should be straight up illegal, hands down. When we are talking about user privacy these details matter and should be regulated. We don't need to regulate presentation elements but we need to regulate equivalency of options and what some defaults should be.

The light grey button with white text makes me smolder, and it's getting a little hotter every time I encounter it.

Better than the white button with white text!

When iOS 7 removed skeumorphism, they made it harder to see controls. Sometimes the controls were white on a white website. Like on Google Chrome address bar! And in YouTube also.

A note for anyone who has trouble telling what's tappable on an iOS device: there's a setting under Accessibility -> Display & Text Size -> Button Shapes that will add outline/fill to some tappable elements, as well as underlines of tappable text

Tesla does the same thing in the latest software.

To configure car settings on the main display, there is a sidebar on the left, and settings to the right.

For example, select "Driving" on the left, and in the pane to the right you can switch the acceleration mode, or steering wheel effort, or regenerative braking level.

If you select "Safety and Security" in the sidebar, there are a number of settings in the pane to the right. The trick? the right pane is scrollable. Like realplayer. Move your finger upwards to reveal a [Data Sharing] button to disable sharing of car data/video clips with Tesla.

Is that the same?

Other panes in that settings screen require scrolling too.

I just checked. There are many options and it's obvious you can scroll on that pane.

The only deceptive thing might be having it at the bottom of the list.

edit: It may be a weird place for that option too thinking about it.

Here's a picture of how it looks by default. I think the sentry mode text and how it's grayed out implies you can scroll.

https://imgur.com/gallery/RqSnWLk

Yeah, that just seems like bad UI more than Dark UI.

Oh, you have a model 3.

On the model S, with the v9 update no other screens were scrollable. I knew you could disable it, and I hunted for quite some time before I found the trick.

I recently got a 2019.x update which added language options to the "Display" pane, and that became the second pane to be scrollable.

(There's also the obvious thought that NO setting should require scrolling to find because you're driving a car)

Ah yes! I knew I actually felt this issue somewhere too and it was on the loaner S. I get it now because of the portrait display. I even double checked before I made my original comment because I was doubting myself.

There are a bunch of little UI issues. Backseat warmers hidden in menus. No physical gloveblock button. You can't press and slide the volume up all the way because you slide off the edge of the screen. Honestly the behind the wheel display on the S was nice. I liked the W/mi gauge better than the white line on the 3 and the full bleed map looks nice.

The decision to get rid of visual scroll indicators is something that I curse nearly every single day.

But it reduces information overload! Improved UX!

I've made so many mistakes because I couldn't tell something overflowed. The scrollbar doesn't have to be big and ugly but it should still be present. At a minimum give a visual indicator that there's more content and then I know to try scrolling to see if it works.

It's a stupid UX fad.

I think even this should be regulated in panes which provide control or information about sensitive user data. If there is overflow, there should be a scroll bar or other visual indicator.

My favorite related example is on my older android phone was that if you told it to not ask for permission for enhanced location again it would grey out the ability to say no.

Maybe my newer android phone has the same unethical behavior ... with that setup it was just a matter of time before I (or someone I handed the phone too) accidentally forever-authorized it.

Not to mention the fact that Google constantly spams you both in your notifications pane and every single Google app for you to turn on location and web history, and even arbitrarily restricts features which would otherwise work without them, such as Google Assistant.

>I hate dark patterns and wish they didn't exist. But I'm not sure that putting the government in charge of policing them is a good idea.

I think after about 30 years of deregulation and "I'm not sure that putting the government in charge of policing them is a good idea", it's time for the pendulum to swing the other way.

Yes that means there may be bad regulations and/or enforcement, but nothing is written in stone. Those things can be fixed as they come up.

Relevant: https://twitter.com/camerondare/status/1143538064402157568

I’m always incredulous about legislation ala carte because lived experience has shown me that the “fix later” part of “legislate first fix later” so rarely happens because of all sorts of forces from political to social mixed with the passage of time.

But maybe I’m a bit bias in my incredulity residing in a state where legislative sessions happen every two years resulting in sometimes two year waits to come back and address a bit of policy that probably could have been deliberated on more in committee.

If Congress legislates they aren't going to be prescribing UI design in the text of the bill. They will give the FTC rulemaking authority and they'll be the cop on the beat that digs into the weeds.

To be fair that isn’t much better in my view, given the FCC’s own behavior recently with dark patterns and public commentary on policy changes that the agency itself promulgated and “requested” public input on

Furthermore I’m not even sure if I agree that regulating UI patterns of web application even falls or should fall to the delegated powers of the FCC; this agency primarily concerns itself with interconnected communications systems like the Public Telephone Switching Network, not social networking sites.

We have a consumer financial protection bureau, if a solution should exist at the Federal level, I’d be far less dubious of a bureau that follows a similar framework as the CFPB to establish guidelines and frameworks for handling user data and disseminating data to the user respectively.

The DETOUR Act empowers the FTC to do this, not the FCC. The FTC already has the responsibility of policing "unfair or deceptive acts or practices" and this is an extension of that.

The bill is 13 pages long. There are affordances for small businesses, this bill would only apply to "large online services" with 100 million users or more within a 30 day period. It seems to be a very thoughtfully written piece of legislation.

https://www.scribd.com/document/405606873/Detour-Act-Final

Ope I misread your agency there. Sorry about that.

Taking a gander at your link before responding further.

Edit: Alright I’ve read the first half, because I’m at lunch and would like to take the time this evening to really sit and pour over it:

I concur with your assessment. This does seem to be well designed and thought out policy, and I could see myself actively supporting it versus my previous position of “no Federal involvement”.

Point, humbly conceded. Thank you for the point of view change.

>But maybe I’m a bit bias

everyone is biased this or that way, what irks me about the "I don't like legislation" comments is that I rarely see a proposed equivalent alternative.

Because just praying that the dark patterns are going away is probably not going to work.

Well then for what it’s worth, I’ve proposed one below that suggests following the framework of the CFPB if Federal government is to be involved at all (I personally think it should not be).

I’m open to your response to that suggestion.

Not everything has an ideal solution. We can't just magically apply regulation, technology or markets and fix every problem.

Sometimes solutions create direct tradeoffs between freedom and security and we need to understand the valuing security can mean a loss of freedom - the PATRIOT act is a prime example.

The best solution is occasionally to just accept that there is no solution, that the tradeoffs aren't worth it and that we should just educate as best we can but accept these things as problems we can't avoid without significant costs.

> Not everything has an ideal solution. We can't just magically apply regulation, technology or markets and fix every problem.

Doing nothing takes us from having "a less than ideal solution" to having "no solution at all".

I think we can try for sensible regulation that strikes a good balance between freedom and security while also acknowledging it might not be a perfect solution in every situation.

The patriot act is a terrible example to use because even before it passed it was clearly going too far in the wrong direction and people concerned with privacy and civil liberties were very vocal in opposing it. The flaws were obvious but included by design and so objections were simply ignored. It simply wasn't an example of government making a good faith effort at sound regulation. It was a blatant power grab which they manipulated frightened people into accepting while ignoring all other objections.

Some of these dark patters are essentially fraud and there absolutely should be a process for reporting them to a governing body for review.

> Doing nothing takes us from having "a less than ideal solution" to having "no solution at all".

No, doing nothing takes us from having a completely harmful and destructive solution that's very difficult to change to having no solution at all.

Sometimes, you need to be willing to accept no solution at all is the least harmful choice.

Well the opposite of that is "do-somethingism". Just because it is an alternative doesn't mean it is better than doing nothing.

To give a snide description - just because you have an infected toenail doesn't mean shooting yourself in the foot would help.

This isn't a "do nothing ever" but "be reasonably sure it would help and be worth the costs".

It is also easier to tell when something would make things far worse than to tell when it could help things.

Yeah, lawmaking has a way of sticking around whether anybody likes it or not. Look how long it's taking to clean up the messes left by the PATRIOT act, or the DMCA... it'll be literally decades, if ever, before all that damage can be undone. It takes so long that people just get used to it as the way things are, and adapt, and then it's even harder to change.

We’re also 40 some odd years from the enacting of FISA, which should be an acronym fresh on the minds of anyone who watches the intersection of technology and regulation as an example of the long tail of legislative scope-creep w/rt NSA programs and activities.

Move fast and break things is bad policy for companies, I don't think it would be any better for the government.

I’m not sure anyone is suggesting that policy, but “do nothing” seems equally bad, if not worse.

The parent was suggesting to regulate first and fix broken regulations later. That's almost literally moving fast and breaking things.

==The parent was suggesting to regulate first==

The assumption is that introducing a bill to regulate dark patterns is the same as "moving fast". This bill hasn't even made it to committee [1] and darkpatterns.org was started in 2010 [2]. I think an argument could be made that the US government is actually moving pretty slowly on this topic. GDPR has been in place for over a year, so there is significant data to understand the implications of this type of regulation.

==and fix broken regulations later.==

How is this different from a software product releasing an updated version? Microsoft Excel is on v16 and I don't think anyone is accusing that product of moving fast and breaking things. I wish we spent more time adding features and fixing bugs from past legislation.

[1] https://www.govtrack.us/congress/bills/116/s1084

[2] https://en.wikipedia.org/wiki/Dark_pattern

To clarify, I wasn't really commenting on the validity on any particular regulation so much as on the idea that "it's time for the pendulum to swing the other way." which seemed to imply (to me, anyway) that it's better to do something, anything, rather than doing what we're doing. I agree that doing effective things is better than nothing, but I think doing ineffective things is worse than nothing.

As to your second point, I also wish we spent more time adding features and fixing bugs from past legislation. I don't think that happens nearly enough, which is a major reason I don't trust our ability to just fix shoddy regulation after the fact.

I took the argument to mean that it is better to regulate bad things in a way that might be ham-fisted, but allow for future improvements than to do nothing about a known problem.

The idea people seem to be repeating is that this will be ineffective. It seems to me that GDPR provides a pretty strong framework to learn from and build off. Europe effectively launched version 1.0, so now we can implement the best parts and eliminate or improve the weakest parts. We aren't starting from scratch or searching in the dark.

“Do nothing” does not appear to be an argument or position being made by anyone present in this portion of the thread.

True, but the lack of a proposed alternative to government regulation is the equivalent of taking that position. Using the CFPB, as you suggested, is still government regulation.

I couldn’t more strongly disagree with that take. Simply not knowing how to address a problem-as an individual-that has been articulated as one in need of solutions is emphatically not the equivalent of taking the position that nothing be done at all or that involved parties abdicate the search for said solutions.

That’s terrible form. I’m sorry. No.

But to your point yes I did propose a similarly frameworked agency like the CFPB, but I’ll also offer that in another comment I had my mind changed on the affair after further discussion with another commenter here about policies already established and powers delegated to already existing Federal agencies.

In reality, we can never know the full impacts of new legislation until it's released into the wild. You said yourself that "legislate first, fix later" makes you incredulous because you feel it is never fixed later.

I am suggesting that the perfect bill which requires no future fixes is non-existent. Thus, advocating for this solution is the equivalent of advocating for nothing. Just as building a product and showing it to nobody is functionally the same as not building a product from the rest-of-the-world's perspective.

You said yourself that "legislate first, fix later" makes you incredulous because you feel it is never fixed later.

Yes. But again, I've already conceded and abandoned this argument after being shown what I thought was good, well crafted policy with definitions, provisions and executions that I am willing to support in the context of this discussion. I happen to believe that all other things being established, context is important. Therefore I don't see the point in trying to refute your assertion of what I've said there, I did take that position, but I've since abandoned it, and conceded the point to a much better position than the one I had previously, therefore what I said earlier about "legislate first, fix later" holds no further relevance to this debate on policy, at least not IMO[1]

I am suggesting that the perfect bill which requires no future fixes is non-existent. Thus, advocating for this solution is the equivalent of advocating for nothing.

You still seem to be arguing against a position of "pass a perfect bill or no bill at all" that no one has made, and has not presented itself in the discussion, so to that note I'm not really sure how to respond.

[1] caveat lector: That being said, I do have strong opinions at large about incrementalism in policy design, but that's for another time and probably another forum.

I would rather the government move fast and break things than let things slide and provide a good argument for keeping things the way they are.

The core idea of 'move fast break things' is fast iteration time. This is actually really good because it means that even when there are issues, they get fixed.

The problem is when you move fast, break things, and then don't fix them. Hell, I'm even OK with being slow to roll things out to make sure you don't break things, just make sure your iteration time period is on the order of weeks or months, not years or decades.

> but nothing is written in stone. Those things can be fixed as they come up.

Unfortunately, this is not true. For example, a right that is taken away will be EXTREMELY difficult to restore again. It's not like a switch that you can flip back and forth until enough evidence is gathered from both states that you know the "correct" way forward... Software has A/B tests for this, law most definitely does not.

Unfortunately regulation does seem to be mostly written in stone. Sure, things can drastically change in events like revolutions but I'm not holding my breath for that (nor do I think that kind of instability is something good to have, it's essentially a huge regression with the opportunity to "redo" things, but as humans we ALWAYS do the same mistakes again sooner or later). How many laws are we re-evaluating every few years and updating/changing them as necessary?

I think if we had almost all laws passed with expiration time it would hopefully result in better/more up to date legislation although the only "temporary" legislation I remember (Patriot Act?) was just passed every time it expired...

30 years of deregulation is a vague, almost meaningless phrase. I'm in pharma and I haven't seen this "deregulation" so to speak. We're seeing more and more regulation surrounding safety of our products (and obviously I agree with that kind of regulation).

>Those things can be fixed as they come up.

I assume "as they come up" means decades later? The government is like a supertanker, changing course is a sloooow process.

Depends how you look at it. Distinguish dark UI pattern from the traditional confidence trick please. The same abuse of trust and guided decision making whilst carefully trying to stay just outside the line marked fraud.

My wife is tech literate, and has put up with my complaining of countless patterns and tech developments over the years, but still managed to get signed up for a subscription when she thought she'd just bought a pair of shoes. A little search showed it a very common issue a few years ago. Maybe it is still.

It's all very well not wanting to make life difficult for startups, but what of the flipping public? I don't think it reasonable to expect people examine every pixel of every page just to avoid getting scammed. The startups, even large ones, are often as bad as the worst snake oil selling SEO and internet marketers.

It's those startups who made regulation increasingly inevitable. We've had 40 years since Thatcher and Reagan started working to make regulation unacceptable and unfashionable. The first 10 years probably did improve things, the next 30 has been too much and caused countless scandals, banking crises and what not.

IF there is no law then how do you fix the problems? I remember years back when a law had to be created so money exchange shops had to display the commission with big fonts under the place they display the prices. The law fixes the issue people had, fixing it with free market and wishful thinking would have cost much more.

I think you can find some good enough rules on how things should be without forcing the use of the exact font family and color.

The EU already applied a similar law with Microsoft's "browser ballot" screen. The requirement is fairness, not any specific presentation detail. Something like "whenever you request permission for something, you must offer the chance to refuse permission equally prominently".

The government can't bring themselves to stop robocalls. I have little if any faith they can save us from something as nebulous and abstract as "dark patterns".

Maybe force websites to be accessible through an API, so if their UI is crap you don't have to use it and you can use a third-party UI instead.

This may be a domain that consumer protection groups could latch on to.

A little bit of lobbying to fix the worst of the problems and shining daylight on the rest might be enough to change the course of enough companies to improve the situation.

The question you are raising is really: Should we build government policy around promoting startups/ new business or protecting consumers?

If it's too subjective for the government to make specific laws then the government should just make vague laws and enforce them.