More fuckin' user interface design
blog.plover.com
blog.plover.com
You can’t A/B test your way through UX/UI in a product like GMail. Sure, it works for a e-commerce front end but just because of UI problems, people aren’t going to ditch the entire GSuite like they would if it was a shopping website. They’d just close it and A/B tests would produce a strong signal.
At Apple, when Steve was around, he had such a keen eye to details. He cared. So if PMs are getting fat checks and not getting scrutinized by their management, I blame the entire ladder upto Pichai for growing such a culture. Engineering at Google is top notch and lots of great things come out of there. Obviously, very smart people for there. I don’t think there are 9 people who don’t get it.
You need a strong dictator at top. Google is becoming a snowflake like organization where it would be impossible to criticize products in favor of not offending anyone - at least that’s what it appears to me. And management doesn’t give a shit about their own products - you're right that the incentives just aren't there.
I’m truly convinced that in order to produce top notch products, you need a military type dictatorship and harsh rejection of bad design - personalities such as Steve Jobs or Elon to make better products. Whether we want to work in such environments is orthogonal and up for debate.
Archive (a box with a label tag on it, https://www.storagegiant.co.uk/wp-content/uploads/a3-archive... from the frontside) Unarchive (same box, but the lid is slightly tilted open)
You know why it works? Because the button also has the text on it that says 'Archive' or 'Retrieve' and the interface actions change accordingly on whether the resource is archived or not.
It's not about iconography, it's about ever increasing loss of context through simplification. It's the curse of the modern UI/UX designer.
But instead Android has most of these buttons on the top of the screen, which is the hardest to reach place on the whole damn thing. And every other app out there uses a completely different set and style of icons and that one curved arrow symbol in the mail app does something completely different from the same button in the note-taking app and the calendar.
It's a convoluted, insane mess.
Even if Microsoft created a Android launcher or fork, it wouldn't really help, because the apps are still free to do whatever they want.
And neither Android nor iOS can compete at text input/editing. It feels like Microsoft put more thought into that component alone than Google does for an entire major Android update.
Yes, it also had a fair share of weird quirks and weaknesses. But in my opinion the Windows Phone 8 UX was easily the best thing Microsoft made since Windows 7, and it keeps being years ahead of Android because Google seemingly rejects any idea that would fix existing issues or reduce confusion.
Change to text, lose weird icons, and gain peace of mind.
- ui consistency across languages and devices
- some languages are longer than others. you can’t just translate the text, this might break the layout. more common at smaller companies without testing resources to cover all the languages available
- mobile has less screen space. going icon only might be necessary. this also goes back to the first point: icon only is easier if it’s icon only everywhere
i’m not saying it’s better. i prefer text
For those who also want to try it in the web interface: Gear->See All Settings->General->Button Labels
Don't forget to scroll all the way to the bottom and save.
In fact, I think that's how it worked before.
As GMail learns from what you mark as spam, that had bad consequences for future emails.
(1) Designing icons that work for a product that's used by billions is really hard. you have to deal not just with how it looks, but how people will see it across different languages, disciplines, experience, and so on. It's practically impossible to do it "right" and to satisfy everyone.
(2) I bet you given #1, the idea here isn't to design an icon that works, but to create a placeholder for a position in the UI, and then train you to click in that general direction. That's why two icons are on the opposite ends of the navigation bar. They know those two are confusing, but they just want you to remember (<- go this way to do X, and go -> that way to do Y).
So obvious in fact, that they removed the text labels. So insanely obvious that it made sense to reduce the size of the icons to like 20px x 20px. So inexplicably obvious that they chose two basically identical icons to convey the message.
So obvious!
These are standard material design icons. Most likely someone needed to add a new button/feature and just picked whatever icon best fit the purpose, or someone designed a new icon that would fit the purpose of whatever feature/button was being added.
With huge apps like Gmail, there's likely just so many different things happening that there's no time allocated to check every small change in a broader context, and these kinds of things slip in and pile up.
Over time, more and more "small" things pile up, and eventually a redesign is necessary because the palette has been polluted with too many things and the software becomes too clunky.
Another option is that no designer even gave this a look, or they saw it too late. This happens often too, designers might be busy working on a big new design and an engineer needs to just put in a button and doesn't think twice about just using one of the icons from the icon library.
Incompetence can explain stuff sometimes, but just like engineering, design is also often a collection of compromises and sometimes a lot of compromises can collude together to form a bigger disaster. The call icon is a good example of this. It's constrained by a lot of things, like overall language rules of material design, and any prior interface design choices made about button styles, toggles, etc. If the designer came up with a totally new style of button, they'd also be scolded by their peers for not maintaining consistency.
All of that said, I'm not making excuses, I'm providing an explanation. I think overall, with any kind of iconography based interactions, when in doubt, add a label. It just works. Icons are abstractions of language and actions, when it's no longer clear what they are abstracting, it's time to use language. It really is that simple.
1. The heuristics by Nielsen https://duckduckgo.com/?q=heuristics+using+Nielsen&t=iphone&...
2. Learn to do usability testing. Fundamentally it's the only way to know what's comprehensible for your target group in a given usage context.
Then there are the HIG human interface guidelines for platforms. There seemed to be a lot to learn from the one of Apple's the last time I checked.
That mitigates some of the damage done by unclear icons. You push the button... oops, that didn't do what I thought it did. Well, now I know, and the Undo button fixed it, so we're all good.
That's not perfect, but it does let regular users and power users coexist without separate modes. The former become the latter smoothly, without having to decide one day "OK, I don't need the screen real estate for labels, so turn them off".
In this case the Undo mode is also very imperfect: GMail has it inconsistently, and it's modal. It exists only in a toast, and not as a general Undo button in one consistent place. It's not a full stack, another thing that Raskin strongly encouraged. That indicates a significant re-think is necessary, not just tweaking the labels.
Anyway... really good book. It's not just UI tips and tricks, but a new way to think about what a UI is.
“What do I want the user to notice? Where am I trying to direct their attention? If I didn’t know what this button does, how do I find out?”
It’s like checking out what you know at the door.
This is MS Outlook's and MacOS's archive button - https://i.imgur.com/Csjpgth.png - if you can't see that, it's a filing cardboard box with it's oversized lid fit on top.
I use MS Outlook web UI and exchange on MacOS with an Apple mail client.
Also then by 'fixing' the archive one to this better one, you sort of also fix the 'Move to Inbox' icon because it's not so similar. The 'Move to Inbox' isn't to bad if it's by itself, what makes it bad as the author says is it's context. I'm not sure what a better 'Move to Inbox' icon would be, I'm not a skilled icon designer.
https://sean-mcbeth.tumblr.com/post/77384411853/gmail-is-a-u...
The TL;DR is: Gmail has a number of modifier-less keyboard shortcuts that can hide messages and dismiss the notification the message has been hidden faster then you can notice if you are typing fast and accidentally lose GUI focus on the message editor area.
Email clients are not video games. The keyboard commands should require modifiers.
gmail2.png still wont load but he's circled it at least in the 2nd image.
And their solution is straight forward; keep a few buttons for the most common actions with an icon along with text indicating what the button does. For further commands, a dropdown is available where all the commands are merely text.
Because usually the problem would have been to design a UI with interface constrains around the size of the text in English, but then »indlæs« (load) turns out to be just a little wider than your box had allocated. But keeping the flexible is the solution (and thus minimise its number).
1. Open Gmail.
2. In the top right, click Settings and then See all settings.
3. Scroll down to the "Button labels" section.
4. Select Text.
5. At the bottom of the page, click Save changes.
The text also has the benefit of making the buttons much bigger / easier to click.
A bookmark still isn't quite robust enough because I still need to scan to find the exact reference when jumping out from the text.
This is why e.g. I read modern hyperlinked PDFs and Wikipedia articles faster and absorb more than their dead-tree equivalents (although I find I read dead-tree equivalents faster for longer linear reads).
And that's just non-interactive, text-based media. Video tutorials for quick how-tos are way way more effective for me than the equivalent book. I haven't had quite as much luck with interactive articles, but there have been a few gems that simply wouldn't be possible with a book (e.g. https://ciechanow.ski/gears/).
Perhaps it's a kind of performance art, and I'm missing the point: that things are harder than we might naively assume at first glance?