The whole gestures thing, besides a few handy ones, requires such a specific set of circumstances (flip the phone over, turn it the right way, make the gesture) to make them workable they are mostly forgotten about.
In general, it's just simpler to press the button.
This is a constant struggle in design, even (especially) when targeting highly technical audiences. We frequently conflate the familiar with the simple or user-friendly. People will shout "how intuitive!" when a designer repeats an interface that's currently familiar in a context that forgets why the original interface was helpful.
So we get a few nice touch gestures: scroll, zoom, etc. We then expand adding more until learning all your gestures (on that particular os for that particular hardware) is more complex than ASL.
The page looks like https://i.imgur.com/RqWJb7Q.png for more than 20 seconds until it finishes loading and then just keeps on downloading more and more and more data (400MB and counting). Pardon me if I ignore their ideas about design.
I think that's unfair. OP is pointing out what he considers to be poor design on a site meant to celebrate great design. Pointing out hypocrisy (not saying I agree with OP or not) is different from unconstructive criticism.
The page design is atrocious. That isn't an accomplishment.
Furthermore, a Google design page where they herald their own designs? Eh...they can pay advertising dollars for that. Google isn't enough of a design leader for such a self-congratulatory pedestal.
When a company is downloading 400MB of a webpage, I don’t care who it is - it needs to be criticized.
Who in the right mind is in charge of this bloat at Google? I wonder what goes into the engineer’s minds when they try to build a page like this one. Goes into the same bucket as IBM Plex font page.
People working at Google are probably given a powerful computer with an outrageously fast Internet connection, which naturally leads to things like this not being noticeable for them. Combine that with a lack of intuitive notions of size[1] and the prevalent dogma of avoiding "premature optimisation" and it's not hard to see how this was created.
[1] It's not just "non-computer people" who don't understand the units. I've taught students who have no trouble writing code or doing the maths in CS, but wouldn't be able to appreciate just how much 1MB of text is, or why trying to send me a 15MB screenshot as an email attachment is a bad idea.
Design.google page scores an appalling, an abysmal score of 11/100.
There is no defense to this behavior - these are professionals who are paid well and are supposed to be the brightest engineers on the planet, and yet they can’t design a page that meets their own standards for quality. 11/100 is shameful.
Spent the whole interview whiteboarding algorithms, forgot to ask about shrinking js bundle sizes. Google doesn’t concern itself with such menial concepts, just too busy saving the world by selling ads.
Avoiding "premature optimization" has become such a ridiculous philosophy after it became mainstream and people forgot what it's actually supposed to mean. We stopped understanding the reasoning, and took those two words and ran with it instead. I frequently have people on IRC/discord ask a bot to tell me about "premature optimization" after I ask a specific question to resolve a bottleneck my profiling has uncovered. It's insane, and so sanctimonious.
I'm a software developer of 3 years and I don't understand why sending a 15MB screenshot as an email attachment is a bad idea?
I agree that intuitive notions of size are tricky. I try to lookup those comparisons about how large the works of Shakespeare or a Tale of Two Cities would be, but it never ever seems to stick. I'm able to remember that movies are roughly 1GB because it's easy and I encounter it a lot as a discrete unit. But yea, songs, text messages, single images, I really struggle with.
In that case it was a 15MB uncompressed BMP of a desktop when the student only needed to attach the files and error messages he was experiencing.
But yea, songs, text messages, single images, I really struggle with.
Those would be MB, <1KB, and KB-MB, respectively. 1MB of text is a lot. In comparison, your comment is 506 bytes, or just slightly smaller than a disk sector (512B). That comment repeated 2000 times would be ~1MB.
The total size if you don't do anything on the page is already bad, but it's also different than if you actually try scrolling. Scroll down to the bottom and watch what happens.
Also lol at them having the gall to give their desktop page a 97/100 and then saying "Use video formats for animated content. Estimated Savings: 42.57s"
And then once you've reached the bottom of the page (and assumed that page is loaded in full) scroll back up. It just keeps loading that data again and again
I love how Google punishes above the folder render blocking in their speed tests, but can't even get it right internally. I've had so many clients complain about render blocking because of Lighthouse, and it take so much time to explain why that isn't always a problem, and sometimes is required when you're using certain js or js libraries.
Supposedly Chrome is testing a splash page for sites that shows average speed test grades, I'm sure they'll whitelist their own products.
To put that into perspective, that's about 37 minutes of uncompressed CD-quality audio, 20 minutes of high-quality 720p H264 video, or a full installation of Windows 98SE. All of which I think offer far more content for the size than this page.
That said, at least the page is still readable with JS and CSS off, unlike a lot of other pages-turned-into-apps I've seen.
Same for me if I don't do anything on the page. But then if I actually scroll the page it gets exuberant. I think it's loading the same assets many times in parallel and possibly repeatedly. Also, "only", lol.
Firefox actually handles this page's bullshit gracefully, by saying "NO! FUCK OFF, WEBPAGE!" ("Will-change memory consumption is too high. Budget limit is the document surface area multiplied by 3 (730032 px). Occurrences of will-change over the budget will be ignored.")
In all fairness, you probably have the "disable cache" box checked in your developer tools. If you turn that off, the multiple duplicate requests in the network tab go away. Presumably the duplicates get served from cache.
Of course it would be better if they didn't make duplicate requests in the first place!
Yeah it works out OK with caching. Very easy to track down with the new initiator tab on the in Chrome Dev Tools. They have _initLightBoxItems which removes all the images/gifs from the dom and then readds them which is bound to the scroll event (with a 50ms throttle).
Not only that, they swap in a fancier font after it loads which has slightly different text sizes. This causes everything to jump around as you're reading it. Wasn't the whole FOUC issue known by designers years and years ago?
It jumped around 2 seconds into load for me, and I'm on reasonably fast WiFi less than 100mi from Google HQ!
>“A Space for Being” explored neuroaesthetics, the field exploring how design impacts our biology and emotions
no arguing about impact, i feel it just by looking at the picture. Being GenX, I'm probably older than the target demography (and the ones who designed it), and that is probably why the impact on me has different sign - it looks to me in a disturbingly unnerving way like a hybrid of prison cell and operating room from dystopias where mind control/altering procedures are performed.
As a creative director at a design [1]firm focussed mainly on B2B products (and a few B2C) it’s fun to be constantly working on the future. It’s hard from my position to look back at 2019. Im too excited about some of the projects we’re helping launch in 2020. I have a hard time with most of these ‘let’s look back at 2019’ collections. Anyone else with me on this?
This seems a lot more like an ad for their products or ideas than a breakdown of great examples of practical designs. I was hoping for examples I could learn from.
A page that talks design in one of the worlds biggest companies - has drawings that looks like they're made from children.
Great, but the reason is obvious. It's to not offend anyone. That is the key of making modern art 2019, design and art that makes even the most grumpy SJWs happy.
Hilariously enough this page would not work in Chrome on Android. Scrolling was totally broken. Firefox Preview worked perfectly though so I got to read it at least.
"The mantra “focus on the user and all else will follow,” is always top of mind here at Google"
Please excuse me if I find that shockingly difficult to believe. Google is one of the most user-hostile of all the large tech companies. I'm genuinely surprised they even had the audacity to write that.
I can only assume you are talking about interacting with Google, the company. I've never succeeded in getting any humans attention in Google, and I suspect that very few people who don't considerable sums with them have managed it. It is indeed like dealing with a gigantic, uninterested, stonewalling, machine.
However, the article wasn't about that. It was about their UI design techniques. When I remember the clusterfuck that was the early versions of Android, I'm truly amazed by how far they have come. Comparing Android Auto to what the car manufacturers provide is literally night and day. The car's native interface is cluttered with distracting detail and a profusion of buttons and controls, whereas the navigation in Android Auto seems stripped down to the bare necessities - just what you need to know. Yet they both have the same feature set, so Google has managed to squeeze the functionality into a UI so simple a single glance at the screen while tells you what you need to know - and not much more.
Google Sheets on Android is to my mind a simply stunning achievement - the pinnacle in UI design. I only discovered it when a desperate situation forced me to use sheets in desperation on the phone. Anyone whose used a speadsheet on a PC knows they support a large variety of ways of interacting with them by dragging, right clicking, highlighting and so on, and despite all that need a large menu at the top for the things not covered. How on earth could you translate that to a tiny screen that can't distinguish between scrolling and dragging? It seemed so hopeless I didn't bother to look for years. Yet when I did look not only did they managed to include almost all of the functionality of the desktop version in a tiny fraction of the space - they made it all discoverable as well.
There is no way this is just luck, or the inspired work of a single man. They had to have employed the engineering techniques like he those described and probably many more besides - rapid iteration, testing many ideas with feed back from multiple groups of users.
Now back to dealing with Google the company. Despite all this good work, Google Sheets (both desktop and Android versions) has a flaw - in some circumstances it can't add up. 1+1=3 is a pretty bad flaw in a spreadsheet. (It's not as simple as that obviously, but it comes up with the wrong result and yet if you export it to say Excel or Libre Office, they get a different, and correct result.) But I don't pay for it, so do you think there is some place I can report this? Absolutely not. They are frustratingly hopeless in some ways.
Google calendar is bad, thought, especially since they removed the ability to hide early morning hours. Nothing in my life happens between 00-08. Why can I not hide it? Because some designer said so.
Now I just use thunderbird. Much better than Gmail/gcal.
I don't know what the graphic style is called but I despise it. How is that type of drawing not something anyone can do? Honestly. Unless these drawings are done by children they are horrible, if they are I apologize and feel bad. I think Box.com or Dropbox used similar graphics early on.. apple had some going with apple.com/today but they have moved along.
I totally agree I find this drawing style of goofy-looking characters with weird proportions that is going through so many tech companies right now so confusing. I hate it.
The "big bodies, tiny heads" style seems exactly the opposite notion of what you'd want to convey with a tech company too[1]. It's like the exact opposite of https://en.wikipedia.org/wiki/Chibi_(slang)
[1] Unless it's some sly reference to "smart technology, dumb people"?
I honestly think it's done to look more "whimsical" and "accessible" and to dodge any sort of accusation of discrimination or exclusion. Which isn't a bad thing at all. Just the art style is so weird and blobby.
It's called pretentious. It's an attempt to appear to be working on a higher level by deliberately discarding aesthetics and what would commonly be thought of as good design. A sort of purposefully ugly. It was a trend in fashion for a while and sadly it looks to have bled over into places like Google in a desire to be hip.
To me it just looks like bad 4chan memes and trying too hard.
What a Google thing to do, punish the internet for having such unoptimized sites but can't even make a good one themselves.
Welcome to the future of the corporate googlenet! Thought my phone froze while scrolling only to realize they locked the UI thread downloading some images.
I was the engineering leader for a grounds-up rebuild of a user dashboard this year. For the people who use it, it's part of how they make their living.
The process of working with design was the most user-hostile experience I've ever seen. There was quite a bit of "users don't know what they want" thrown around when people saw the beta and complained. It was used quite selectively, geared towards things the designer disagreed with. When the dashboard left beta, it was universally panned. All our KPIs tanked. What's more, it was considered a success because "redesigns are always poorly received." At all the jobs I've worked at, I've yet to discern any true rules that designers follow that aren't rationalizations.
I bring this up because nothing about this site...from its design to its content...resonates with me. If this is truly the best of 2019 then designers REALLY need to take a long, hard look at their culture.
Well, it’s stuff that Google did in 2019 that it wanted to highlight, so take "best of" with a grain of salt.
I’d be interested in a similar list from an independent perspective that doesn’t skew so high concept. Any improved login flows, transit ticketing experiences, accessibility wins, etc?
If memory serves, this app used to have a 4.5+ rating, now it's down to 3.2.
I did find the old 5.10.1 version of the app with a search for "download wunderground android", so I installed that and have a usable app again. Until they make some server change that breaks it.
I have had a similar issue with a UX/design person. I think the main problem though is process. The design person should be challenged and accept criticism, and revisit their designs. Engineers can also fall into similar traps of following trends and over-engineering things (smart people like to make complex things without even realizing it). Without the structure in place to correct these oversteps, designers can have similar problems.
The interesting thing for me isn't that some people use motivated reasoning to push through changes for which they have a personal preference, but that the people who do this sometimes manage to gain power within organizations and insulation from consequences. You see the same pattern in engineering sometimes.
UX Designer (at Google) here. I feel your pain. For any design to work, it needs to be grounded in a firm understanding of the user's behavior. I agree that design intuition is limited by one's own view of the world, so it's critical to gather the right data before going further than speculative designs. This can be accomplished through a variety of means, but observation and qualitative interviews prior to "design" are usually effective.
The design should be validated with different users from the same cohort prior to building it. Functional prototypes work well here. Ignoring user feedback here is a huge mistake on the part of the designer. Design is relatively cheap, and it's easy to iterate and check your assumptions with users before anything is set in stone. Correct me if I'm wrong, but this step would have increased trust with your engineering team. Often due to ignorance, some designers don't appreciate how much effort goes into these builds, and how early architectural decisions shape the product in ways that make it difficult to change later.
Acknowledged, that this is a "perfect world" scenario and business pressure might require augmenting the process. An experienced designer would understand how to do this and still get meaningful results.
71 comments
[ 5.5 ms ] story [ 157 ms ] threadIn general, it's just simpler to press the button.
So we get a few nice touch gestures: scroll, zoom, etc. We then expand adding more until learning all your gestures (on that particular os for that particular hardware) is more complex than ASL.
Furthermore, a Google design page where they herald their own designs? Eh...they can pay advertising dollars for that. Google isn't enough of a design leader for such a self-congratulatory pedestal.
When a company is downloading 400MB of a webpage, I don’t care who it is - it needs to be criticized.
Who in the right mind is in charge of this bloat at Google? I wonder what goes into the engineer’s minds when they try to build a page like this one. Goes into the same bucket as IBM Plex font page.
[1] It's not just "non-computer people" who don't understand the units. I've taught students who have no trouble writing code or doing the maths in CS, but wouldn't be able to appreciate just how much 1MB of text is, or why trying to send me a 15MB screenshot as an email attachment is a bad idea.
I presume engineers at Google don’t run their own performance/speed utility: https://developers.google.com/speed/pagespeed/insights/?url=...
Design.google page scores an appalling, an abysmal score of 11/100.
There is no defense to this behavior - these are professionals who are paid well and are supposed to be the brightest engineers on the planet, and yet they can’t design a page that meets their own standards for quality. 11/100 is shameful.
I agree that intuitive notions of size are tricky. I try to lookup those comparisons about how large the works of Shakespeare or a Tale of Two Cities would be, but it never ever seems to stick. I'm able to remember that movies are roughly 1GB because it's easy and I encounter it a lot as a discrete unit. But yea, songs, text messages, single images, I really struggle with.
But yea, songs, text messages, single images, I really struggle with.
Those would be MB, <1KB, and KB-MB, respectively. 1MB of text is a lot. In comparison, your comment is 506 bytes, or just slightly smaller than a disk sector (512B). That comment repeated 2000 times would be ~1MB.
edit: so I think ive settled on a good rule of thumb: movie: ~3GB song: ~3MB page of text: ~3KB
15 MB is also absurdly wasteful. It would probably be <1 MB in a compressed format like JPG or PNG.
> Avoid enormous network payloads Total size was 64,215 KB
Also lol at them having the gall to give their desktop page a 97/100 and then saying "Use video formats for animated content. Estimated Savings: 42.57s"
42.57s!
Supposedly Chrome is testing a splash page for sites that shows average speed test grades, I'm sure they'll whitelist their own products.
To put that into perspective, that's about 37 minutes of uncompressed CD-quality audio, 20 minutes of high-quality 720p H264 video, or a full installation of Windows 98SE. All of which I think offer far more content for the size than this page.
That said, at least the page is still readable with JS and CSS off, unlike a lot of other pages-turned-into-apps I've seen.
Look at this hilarious bucket of shit: https://i.imgur.com/36CBn5y.png And that's in Chrome!
Firefox actually handles this page's bullshit gracefully, by saying "NO! FUCK OFF, WEBPAGE!" ("Will-change memory consumption is too high. Budget limit is the document surface area multiplied by 3 (730032 px). Occurrences of will-change over the budget will be ignored.")
Of course it would be better if they didn't make duplicate requests in the first place!
It jumped around 2 seconds into load for me, and I'm on reasonably fast WiFi less than 100mi from Google HQ!
no arguing about impact, i feel it just by looking at the picture. Being GenX, I'm probably older than the target demography (and the ones who designed it), and that is probably why the impact on me has different sign - it looks to me in a disturbingly unnerving way like a hybrid of prison cell and operating room from dystopias where mind control/altering procedures are performed.
[1] http://fairpixels.pro
Combined with the color scheme (and as others have noted, page performance), I'd vote this for worst design of 2019.
Great, but the reason is obvious. It's to not offend anyone. That is the key of making modern art 2019, design and art that makes even the most grumpy SJWs happy.
More seriously, some of those drawings seem more like they belong in an article about obesity.
Self-righteous, convinced to be the best. A page full of awards to themselves that barely loads making it de facto the worst design ever.
"The mantra “focus on the user and all else will follow,” is always top of mind here at Google"
Please excuse me if I find that shockingly difficult to believe. Google is one of the most user-hostile of all the large tech companies. I'm genuinely surprised they even had the audacity to write that.
However, the article wasn't about that. It was about their UI design techniques. When I remember the clusterfuck that was the early versions of Android, I'm truly amazed by how far they have come. Comparing Android Auto to what the car manufacturers provide is literally night and day. The car's native interface is cluttered with distracting detail and a profusion of buttons and controls, whereas the navigation in Android Auto seems stripped down to the bare necessities - just what you need to know. Yet they both have the same feature set, so Google has managed to squeeze the functionality into a UI so simple a single glance at the screen while tells you what you need to know - and not much more.
Google Sheets on Android is to my mind a simply stunning achievement - the pinnacle in UI design. I only discovered it when a desperate situation forced me to use sheets in desperation on the phone. Anyone whose used a speadsheet on a PC knows they support a large variety of ways of interacting with them by dragging, right clicking, highlighting and so on, and despite all that need a large menu at the top for the things not covered. How on earth could you translate that to a tiny screen that can't distinguish between scrolling and dragging? It seemed so hopeless I didn't bother to look for years. Yet when I did look not only did they managed to include almost all of the functionality of the desktop version in a tiny fraction of the space - they made it all discoverable as well.
There is no way this is just luck, or the inspired work of a single man. They had to have employed the engineering techniques like he those described and probably many more besides - rapid iteration, testing many ideas with feed back from multiple groups of users.
Now back to dealing with Google the company. Despite all this good work, Google Sheets (both desktop and Android versions) has a flaw - in some circumstances it can't add up. 1+1=3 is a pretty bad flaw in a spreadsheet. (It's not as simple as that obviously, but it comes up with the wrong result and yet if you export it to say Excel or Libre Office, they get a different, and correct result.) But I don't pay for it, so do you think there is some place I can report this? Absolutely not. They are frustratingly hopeless in some ways.
Now I just use thunderbird. Much better than Gmail/gcal.
Maybe some of the HN Google employees can explain why AMP wasn’t used for this slow page?
[1] Unless it's some sly reference to "smart technology, dumb people"?
To me it just looks like bad 4chan memes and trying too hard.
Welcome to the future of the corporate googlenet! Thought my phone froze while scrolling only to realize they locked the UI thread downloading some images.
The process of working with design was the most user-hostile experience I've ever seen. There was quite a bit of "users don't know what they want" thrown around when people saw the beta and complained. It was used quite selectively, geared towards things the designer disagreed with. When the dashboard left beta, it was universally panned. All our KPIs tanked. What's more, it was considered a success because "redesigns are always poorly received." At all the jobs I've worked at, I've yet to discern any true rules that designers follow that aren't rationalizations.
I bring this up because nothing about this site...from its design to its content...resonates with me. If this is truly the best of 2019 then designers REALLY need to take a long, hard look at their culture.
I’d be interested in a similar list from an independent perspective that doesn’t skew so high concept. Any improved login flows, transit ticketing experiences, accessibility wins, etc?
https://play.google.com/store/apps/details?id=com.wundergrou...
If memory serves, this app used to have a 4.5+ rating, now it's down to 3.2.
I did find the old 5.10.1 version of the app with a search for "download wunderground android", so I installed that and have a usable app again. Until they make some server change that breaks it.
The design should be validated with different users from the same cohort prior to building it. Functional prototypes work well here. Ignoring user feedback here is a huge mistake on the part of the designer. Design is relatively cheap, and it's easy to iterate and check your assumptions with users before anything is set in stone. Correct me if I'm wrong, but this step would have increased trust with your engineering team. Often due to ignorance, some designers don't appreciate how much effort goes into these builds, and how early architectural decisions shape the product in ways that make it difficult to change later.
Acknowledged, that this is a "perfect world" scenario and business pressure might require augmenting the process. An experienced designer would understand how to do this and still get meaningful results.