Hacker News Reader

Many of the things you describe, describe my mother to a tee.

She uses an iPad, but progressively Apple has complicated the interface to fit in more and more features, which has made it harder and harder for her to use the device simply.

---

One recent example is the Notes app, which today includes folders, tags, and many other fancy features. I only discovered recently that she had still been using the app as though she did the day she started using Notes, which must've been c. iOS 6. She didn't even notice the top-left "< Folders" button.

Apple's design language is "Clean" and "Modern", but often fails to convey to non-technical users that a even a button is...a button, since iOS/iPadOS use colored text with no background to convey that something is a button.

---

One other example was a mistake I made while configuring a web browser for her. I absent-mindedly turned on "Hide panels automatically", and it wasn't until I figured out what my mother meant when she complained "SOMETHING POPS OUT AND THEN JUST VANISHES" of what I had done wrong.

My pet peeves are icons. I am sick and tired of button actions being communicated with a meaningless set of pictograms I have never seen before. Give. Me. Text.

I'm with you there. Having spent so much time on the command line, it's so much more straightforward to just have a label that plainly describes what something is or does.

As I understand it, imagery is (ideally supposed to be) language agnostic, like the "icons" for events at the Olympics, though it doesn't always work out as intended.

It's unfortunate that designers have since removed all textures/colors/depths from icons, leaving us with basic lines and shapes, and the result isn't pleasant to look at and very tough to recognize.

The idea is, that text is great when you come to the UI for the first time. But, as you use the software every day, it's slow to read the text, so your eye becomes accustomed to the icons and you can use them as a shortcut to save yourself having to read the text. The icons should all be different shapes (as the eye sees the outline of a shape the quickest), be different color (but not only different colors as some people are color blind). I think this used to be common UI knowledge but I remember reading it in design guidelines for the original Mac OS X 10.0.

Somehow this has all gone wrong through: no text so the first time you use it you've no idea what the icons mean, they are all the same color and pretty much the same shape so you don't even get the speed of recognition.

And all of these problems are well documented, researched and have viable solutions (if anyone from Apple needs help, hire me :)

So either the design team is ignoring 20 plus years of research or they just don’t know what they are doing.

My go to design manifesto is making any screen idiot proof. Mistakes can always be made, but you can safely back out from it and you always understand the consequences (aka you learn)

> So either the design team is ignoring 20 plus years of research or they just don’t know what they are doing

The design team has been ignoring Apple’s own HCI guidelines for years now

This is why they published new ones this year. Samples floating on the internets seem to hint at lack of concrete examples, though, and in macOS Ventura they reportedly don’t even adhere to the new guidelines.

Guidelines for thee, but not for me, I guess.

I would love to see the workflow of the design team offices and how they work.

I carefully coached my husband's 85 year old aunt in how to look at the pictures she took on her iPhone and then send the ones she wants printed to our shared album. She had the steps meticulously written out. She was good to go!

Then Apple changed the UI a little bit. I can't begin to explain it to her over the phone in my second language, because for starters, I don't have a non-updated device to be able to compare and find the exact difference. I felt a bit of a jar and was able to roll with it, but couldn't tell you what the difference is exactly.

LTS (guaranteed non-change for existing features, at the cost of possibly not getting all new features) for UI would be amazing.

And do not get me started on the drama involved in an iOS major version update and the requisite iCloud password entry to even start the phone. She has the password written down, but since it's helpfully hidden as she types, she can't see that she was a letter off or missed a shift. I ended up sending her to the Vodafone corporate store with her password and 3 reset answers to have one of them type it in for her, as we're on the other side of the country, and she kind of needed to be able to make and take phone calls.

Stylish has clearly taken priority over functional in mainstream tech design.

I just got a new Android phone that iterated 1 version in the operating system and product. The old phone had a persistent triangle symbol to press for "back". Now that symbol is gone and I'm supposed to swipe from the edge of the screen for the same "back" function. Except, lots of apps use a similar motion for other functions, including Firefox for closing tabs. So, now I'm constantly swiping a few milimeters too close to the screen edge and going "back" when I want to close a tab, with few clear visual clues as to what is going wrong. This is just 1 of many UI fails on the new phone, very frustrating for me and I would guess extremely and unnecessarily confusing for "normal" people.

My Samsung XCover has physical keys for back/home/switch task and that'll be a hard requirement for any future phones I'll even consider. Can't patch those out without really wanting to fuck me over.

I suppose outdated features like good grip and removable battery are a bit much to ask for.

Stylish sells better? Or why does it happen?

People don't have time to dig into usability details, instead they get an app because it gives a nice looking first impression? Could "rewards" in a company work in similar ways -- impressing managers with sth nice looking?

This is why I switched to apple and never looked back, can’t stand OSes manipulating interface controls arbitrarily after updates.

> She had the steps meticulously written out. She was good to go!

On multiple occasions when explaining computer stuff, I explicitely said something like "don't write anything down, but try to understand the metaphor", because the software is a moving target. I've been also asked about "a manual for this computer", when they actually wanted a printed manual that explains all the software on the computer.

Modern day electronics aren't really made for most people who are born before about 1960. Understanding all metaphors intuitively probably requires having been exposed early in life.

> She has the password written down, but since it's helpfully hidden as she types, she can't see that she was a letter off or missed a shift.

If someone had to design a process to mock and torture the elderly trying to use modern devices, that's the one. Along with messages that time out and unexpected events that steal focus.

It’s not just a “old” people problem… unless I’m old, 40 does feel old.

The UI on most mobile devices are terrible in one way or another. There are zero discoverability, things will hide, lock, block or disappear with no clear way of getting back.

Often I feel it’s easier to go to the computer. As an example: I cannot use the banking app on the phone anymore. In an attempt to make it easier to use, they hid/removed the features I used, such as “how much is in my account right now?” or “pay a bill”. That stuff is now hidden in a complex menu system. I can manage the MasterCard I don’t have though.

Designing good UI is an incredible skill, one that few people gave, and one even fewer are willing to pay for.

> Designing good UI is an incredible skill, one that few people gave, and one even fewer are willing to pay for.

I don't think it's even just that. If any of these companies did usability tests they'd discover quickly enough that their designs don't work. I guess they just don't bother.

I have worked in software development since the mid 90s, often freelance so I've probably been at 20-30 companies in total. All of these companies produced software that had some sort of UI. I think only a single time was any usability testing done, i.e. get a normal person sat in front of the software and get them to perform a task and see if they can figure it out on their own. Normally the process is that the designer asserts that this particular way is best, and that's the way it gets implemented and that's what the customer gets to use.

Reading Alan Cooper's "About Face: The Essentials of Interaction Design" made a big impression on me; UI work is really fascinating and rewarding, but it is honoured almost exclusively in the breach thesedays. I can understand companies knocking up intranet based applications for limited internal use cost-cutting, but in many cases these are massive corporations rolling out atrocious UIs to the public.

Part of the problem imo is that everyone thinks "how hard can it be?" and that they have good ideas for UI design. They therefore aren't interested in reading Cooper or submitting their brainwaves to even peer scrutiny.

> honoured ... in the breach

What does that mean? (Not a native speaker)

Don't Make Me Think is another good book I think :-)

Just means it's mostly ignored. And yes, it is a great book.

I'm not defending Apple, but tech-oriented users have always been pushing for more features.

The question is should there be a middle ground between keeping the product trivially simple, or cramming an increasing amount of features to satisfy power users. Either extreme works well for a specific subset of Apple's user base; a middle ground is more prone to leaving everybody confused and/or unsatisfied.

I have zero UI design experience so can't really chime in. Maybe a toggle option inside a menu that swaps app modes? Leave it off and you have a very basic Notes app; toggle it and you get all the bells and whistles, tags, folders, parsing... But this totally causes the problem you describe in your last paragraph.

There's two problems.

Many UIs today are objectively bad and unintuitive (like the Gmail example in the article). UIs should always strive to eliminate the "don't know what you don't know" quadrant: common sense should be enough to navigate the system safely. Perfect examples are the Gmail example in the article, or the "hanging up the phone doesn't hang up the phone" problem of landlines[0] where you can do "everything right" and still get scammed. The judgment metric here is whether a decently intelligent people could fall victims to these; they can, so they're objectively bad UIs. They violate common sense.

Then there's another group, that severely struggles with computing abstractions. (These can be perfectly smart people in other areas). They can have slightly degraded motor skills and end up swiping or tapping inadvertently and getting lost. The kind of people that thinks a locked phone is "turned off". That get confused when Control Center or Notification Center show up, because the swipe was inadvertent and they don't understand the abstraction ("where did my video go?").

It is utterly impossible to design the same UI for this group of people and for everyone else. For them, you'd ideally permanently dedicate a fifth of the screen to a never-changing menu, with common operations (copy, paste, switch app, see notifications, check email, find app). For most people, menu bars would be terrible on phones and tablets since they'd need to be always visible, and big enough to be tappable, and would waste collosal screen space; but some people would prefer that.

The UI would need to have a permanent banner that explicitly labels the UI context, essentially a "permanent tutorial mode" that explains computing abstractions, for example with a banner that says "YOU ARE CURRENTLY VIEWING RECENT NOTIFICATIONS. [GO BACK]" (otherwise, you get inadvertent swipes, and "hey, where'd my video go!").

Designing "one UI for everybody" is hopeless. This is an obvious accessibility issue; and I don't understand why no company is building optional "tutorial modes" into their UIs since I'd assume the amount of people struggling with computers is a lot higher than the amount of blind or paralyzed people.

[0]: https://bc.ctvnews.ca/beware-of-the-delayed-disconnect-phone...

On iOS you can change app settings outside of the app itself. Why not default to simple, and allow technical users to enable the more complex features they want by going through the Settings app?

When Mom was still alive, I installed the tvOS beta profile on her iPad mini 4 to prevent it from receiving updates, so that the interface wouldn't change out from under her (she'd already transitioned from Android to iOS and I didn't want to have to keep teaching her the new UI changes). Wouldn't have been so bad if we had the ability to install whatever version of iOS we wanted.

She is not deaf or disabled nor visually impaired and yet the complexity of using this device presents serious obstacles.

I'm an experienced developer in the software industry and some of the stuff you mention still catches me off-guard --- especially the whole "disappearing controls" thing. https://news.ycombinator.com/item?id=24965293

IMO controls that require some kind of time-related action (e.g., magically disappearing popups and buttons that hide themselves) should violate accessibility standards.

Agreed; I think games --- especially ones where realtime reaction is required --- are the only acceptable use of such controls. Otherwise they're just a really unpleasant "race condition".

Heck, I'm a lifelong-computer-obsessed software engineer in my 30's and almost all those bullet points have an adverse or frustrating effect for me as well. I am "purposely lazy" and spend as little effort as possible deciphering user interfaces. There's simply zero excuse for non-intuitive software anymore. I refuse to expend my valuable time and cognitive energy making up for the shortcomings of a software vendor, and it regularly surprises/disappoints me just how much effort people expend in that way.

In my opinion, it's practically an insult to provide software or a "service" that essentially demands the person interacting with it slog through countless arcane or obtuse interfaces which could have been totally simple, concise and straightforward. There have been literal decades of prior art on subtly nudging a user through the most obvious and "least surprising" ways to complete their intended task.

If Google Duo stops working out, we've had great experience using meet.jit.si with my wife's great-aunt. She's in her mid-80s, and English is her second language. We set up a room months ago and sent her the link, and each time we talk weekly she goes to the same place, clicks the same link, and the whole thing just works for her. It was simple enough that we were able to walk through setup over the phone (we're on the West Coast, the great aunt is in Germany).

No ads (just checked with an adblocker turned off), and the same link always dumps you in the same always-active room. As a bonus the whole thing is open source.

No, I don't work for them or have any interest whatsoever other than being a happy user.

I also tried a setup with meet.jit.si for my parents. As you did, I set up a 'permanent' room where everyone in the extended family could connect at any time. Then I set up a Raspberry Pi connected to one HDMI port on their TV, and I used one of those Logitech cams with a built-in mic which I placed near the TV, and connected to another HDMI port on the Pi. I set up Chromium to start automatically, in kiosk mode, and to go directly to the Jitsi room.

Finally I used a little relay board connected to a couple of GPIO pins on the Pi and programmed that to turn off and on the Pi. The relay board was remote controlled (infrared, worked fine). I used a small, but not too small, remote control with one single big red button. When a scheduled or agreed (maybe by a phone call) meeting was due, my father pressed the button and the TV switched to the Pi HDMI (turning on the TV if it wasn't already on), showing the Jitsi room. With maybe grandchildren and great-grandchildren, and my parents could just sit in their chairs and watch and talk via the TV. When finished my father pressed the button again, and I had programmed the Pi, as part of its shutdown sequence, to switch the TV to the other, normal HDMI port. It's possible to command the TV to switch inputs, via commands to HDMI, unfortunately it's not possible to switch to antenna input but my parents were on cable and used an HDMI port for that. So with this setup they would go from the Jitsi room to whatever they were watching before. Though the TV would be left on even if it was off before they connected - I didn't find a way to control the TV that detailed.

This worked well, as far as it went, the issue I had was that it wasn't practically possible to set up a power supply for that little relay board so I used a couple of 9V batteries instead. But it turned out that the board would suck the batteries empty in no time. So, as a prototype, it worked, but I would have to make a proper power supply for that to be fully useful. After a few months my parents had to move to a care home so the project was put on a shelf.

That sounds so fantastic. This would be a great productd for a startup. Just a small box, USB-C powered, which performs these easy simple tasks, but absolutely reliable and a pleasure to use. Perhaps like 4 buttons, nicely differently colored or labeled, for several rooms. I am thinking, people had never troubles operating their TVs when they just had buttons for a few stations and the volume.

Even as a software delveloper, I would be delighted to use that, for the elegance of the concept and not having to dig through menues, toolbars and whatever else.

(Too late to edit - a small error there. The Logitech cam was of course connected to a USB port on the Pi. The two HDMI connections in question were two ports on the TV, one with the Pi as source, the other with the normal cable box as input)

> - zoom audio needed active confirmation to connect to audio.

I don't know if it's just me, but even as an "IT person", I never understood this "connect to audio" phrase, I just learnt that it is required for calls...

It's there to protect users. They need to give consent to open the audio channel. As a usability / ux person, this one feature I actually like.

Protect them from what? Being able to listen to the call?

To protect them from accidentally speaking their mind, mentioning things that might get them in trouble if the other side hears it. to prevent this, Zoom (and others) ask you to click to open the mic. It's a GDPR thing: The default setting in most tools is that the user has to actively switch on their camera and microphone when they begin to participate.

Auto-muting new users is not the issue (every reasonable video chat software does this). The problem is that in Zoom you have to "join audio" before hearing anything. I guess they do this since you could in principle also join the call by phone. Anyway, this behavior is extremely confusing for inexperienced users.

Ah, thanks for clarifying, I completely agree.

> you have to "join audio" before hearing anything

If it's a private meeting, you're going to listen only, but

you're at a public place and didn't yet plug in your headphones?

I could add a lot more items and annoyances on my own. That makes me wonder, why don't we have some global list of iOS, Android, Macos, Windows and Browser usability issues?

>zoom audio needed active confirmation to connect to audio. On a tablet, where no options are available but the over-wifi audio.

It is annoying they ask it every time but there is, in fact, a different option: option to call your phone (which most people have) which could have lower latency and higher reliability. I have succesfully used it in places where internet is spotty but adding an extra step to join a call is too much.

"I am to not trusting anything online and presuming the worst by default" ... It's exactly the same IRL ...

That's what happens when IT companies hire people that know how to invert a binary tree but know jack crap about UI/UX and human interaction (or you know, about just what it takes to make a call)

Also Zoom on PC is usable, Zoom on phones is awful

Someone failed a whiteboard interview?

Those people (software engineers) aren’t making the decisions around the product in larger companies. These are done by Product Managers and UX Specialists/Designers.

They (mostly) don't make the bigger decisions, bigger design choices, true

But for minor stuff/fixes it's easy to slip something that's awkward or not very usable.

Or a legacy interface is still being used (because the developer thought it out - since the company was smaller, or a specific screen was overlooked) and that's not a great interface

That may be the case in companies without proper responsibility distinction but in those with (usually the bigger ones), not even small decisions are made solely by software engineers. They are always double checked by the Product Management because any requested change has to come from the Product Management and has to be signed-off by them. In a company like Zoom everything you are seeing in the UI isn’t the result of something that slipped through solely by the engineer but was reviewed and thought out by multiple people who don’t touch the code.

So I can be mad at more than one person, when Google Drive's or iOS' user interface is misbehaving