Who Killed the Scrollbar? | Hacker News Reader

Who Killed the Scrollbar?

ello.co

351 pointsdredmorbius316 comments

Flipping flip i agree! It's not just scrollbars. I hate the mobilification of everything. I hate that desktop designs are imitating the nonexistent information density of phones. Bring back menubars, and extensive left-hand hierarchies . Even in mobile, it's more practical for me. I like a full page where i can zoom in and out faster between its parts rather than scrolling up and down all the time (I browse reddit in old mode - it's just better)

Incessant hamburgerization is something that I also find annoying. Any time I have to set up Firefox, the very first thing I do is enable the menu bar.

My desktop isn't a tablet - stop trying to make it one!

I rarely use the menu bar so I just hit alt when I need it. I'd bet lots of users don't even know it exists though. Especially with Chrome hiding so many of their options

I can see why they did it. Now that screens are ~universally 16:9, vertical space is at a premium. It would be nice if the hamburger menu actually replicated the same functionality though.

I think that if vertical space were indeed the reason, then we wouldn't have huge, touch-sized widgets everywhere and 32-pixel titlebars. There are applications where simply reducing the padding on most widgets to sane values wins you 3-4 menubars' worth of space.

I'm sure vertical space is the most often-repeated argument, but I'm skeptical that it's indeed the reason behind this trend.

Firefox in particular is themable; mine is pretty compact.

16:9 is a ratio whereas the amount of vertical space is an absolute amount. I don't think the two are related, unless screens got wider at the expense of height, which surely is not the case.

They did lose height; 1600×1200 → 1920×1080. For a while you could find 1920×1200 if you looked hard and spent more, but at the next increment everything is 16:9.

I have 2 1920x1200 and there is no reasonable upgrade except maybe something like 40" 4K (and not in retina/high dpi mode). 2560x1600 would be ok, but it's too expensive.

3440x1440. Less pixels overall than your current setup, but incredibly nice to use. Also affordable, you can get them for under $400 during sales / end of year clearance.

Seriously, for coding I absolutely love a 1920x1080 screen rotated 90°. Still wide enough to see an entire (sane) line, but more vertical space is just wonderful. Can't watch video, obviously, but for a work computer the vertical space is the absolute best. Improves web browsing as well for most sites.

I tried a 36", I can't stand it as a display. Fine for a TV, but it's a little much for a monitor. Although I can't stand more than 2 displays, so YMMV.

I went to 3840x1600 - as wide as two Full HD screens but with a bit more vertical space and no bevel in the middle.

They did get wider at the expense of height. Monitors are sold by diagonal size, and a 15" 4:3 monitor is 9 inches tall while a 15" 16:9 one is only 7.4" tall.

Agree! "Mobile First" design thinking should not mean "Mobile Only." What was supposed to be a call to consider all of the devices your users may be utilizing, it just transferred the preference from one type of device to another.

Why can't we just have sites/apps/SPAs/whatever that look and work well on whatever device they supposedly support?

> Why can't we just have sites/apps/SPAs/whatever that look and work well on whatever device they supposedly support?

Indeed. There seems to be a weird idea that it's possible to make a UI that works well on all form factors. I don't see how that's possible, and I've never seen it happen in practice.

What actually happens is that you have a UI that works well on one form factor and badly on the others, or you have a UI that works poorly on them all.

I think NYTimes.com is a good example. Theres a hamburger menu that expands to the whole screen when on mobile or tablet; its a sidebar on desktop. The whole arrangement of articles becomes linear on mobile; it isnt on desktop. It looks like someone figured out how it should look at every state, and then developed a way to gracefully switch between them -- not that it was a mobile UI that someone decided to enlarge on desktop.

This is why we have media queries in CSS!

It's interesting that you turn on the menu bar. To me this isn't a "tablet" feature at all--it's just a "screen real estate is precious" feature. Even when hidden it's only an Alt press away.

This must be a Linux/windows paradigm? I am lost looking at Firefox on my Mac.

Edit: apparently on PCs the per-window menu is activated with Alt, TIL

On Macs the menu bar is intact, same place it ever was, detached from the firefox window at the top of the screen. The hamburger menu on macos is basically redundant.

I always turn on the menu bar on the desktop. Screen real estate is valuable, but using some on a menu bar is a very worthwhile expenditure, in my opinion.

I know about the "Alt" key, but I find the need to press that to pop up the menu to be incredibly annoying.

I'm super curious--how often do you actually use the menu and what do you use it for? I ask because I literally can't remember the last time I did in Firefox; everything's a keyboard shortcut. Fullscreen is F11, inspector is F12, refresh is F5 (even go-to-URL-bar is F6).

go-to-URL-bar is also Ctrl+L which I tend to favor, for some reason

Also I miss typing ".." in the URL bar to go to my home page (I'll admit this was an iexplore quirk, albeit a really convenient one...)

I pull down the History menu reasonably often. I'm one of the few people who still uses bookmarks too. I found that hitting alt was annoying and I'd forget about the menu and try to find stuff on the Hamburger menu instead and get annoyed.

Huh. I use bookmarks a lot, too--but I use them via the omnibar. Type the first few letters and I'll find it from there.

Interesting. Thanks for sharing.

Now that you mention it, I also use the history menu pretty frequently. I don't use the bookmark facility in browsers, though, because I run my own bookmark server (so I have access to my bookmarks no matter what computer or browser I'm using).

Maybe you know this already, but Firefox has shortcuts for both of those. History is Ctrl+H and bookmarks is Ctrl+B. Ctrl+Shift+H and Ctrl+Shift+B to get them in a separate window.

The Ctrl-H history opens up the annoying sidebar though, and then you have to click to expand. It's not as ergonomic as the menu. The Bookmarks have the same problem.

Just put the History menu dropdown in the toolbar beside the hamburger menu.

> I'm super curious--how often do you actually use the menu and what do you use it for?

I use the menu really frequently -- for pretty much anything that isn't a toolbar button.

I tend not to use keyboard shortcuts generally, but particularly if I'm just reading a website -- in that situation, my hands are on the mouse, not the keyboard, and moving to the keyboard to type a shortcut is more inconvenient than just being able to use two clicks on a menu.

The screen real estate isn't particularly precious on a desktop / laptop with a large, high-resolution screen (much less precious than mobile, for sure), but I also like to get rid of menu bars, tool bars, etc. just because they're visual clutter, and like you say, a keyboard makes up for them.

Ironically, I do enable them on my tablet, because I don't always have a keyboard attached to it. Tablets generally have enough space for full menus / toolbars, and lacking a keyboard they're arguably the form factor most in need of them.

I use three 1440p 27" monitors and vertical real estate is still something I'd seriously inconvenience an adorable puppy for more of. If my desk was a little better-shaped for it and I didn't need a place to put speakers, etc., I'd go back to using at least one vertical display.

What I really wish for are 2000x2000-ish square screens, ~20 inch diagonal with minimal bezel, so that I can stack 3x2 of them and get to work at last instead of flipping wasting screen space and head movements on overly wide screens.

Vertical bars would have enough space.

We've got bigger higher resolution screens for every device, screen real estate is only getting less precious and we're still losing affordances like scroll bars and being able to see the top level of the menu hierarchy.

Not only that but a big pitch of the tablet itself is to get the desktop web browsing experience.

Good god, and I'm over here living like a plebe thinking I had to just deal with the hamburger.

Thank you to the poster below for the "alt" shortcut and for reminding me I can change the settings!

I don't have a problem with the concept of the hamburger except what they put in it is less useful than the menu bar.

There is a style of roundabout called a hamburger roundabout. It's controlled by lights, and has a two-way lane running straight through the middle.

I don't drive much, these days; but even as a passenger, they do my head in. No More Hamburgers! Boo!

> My desktop isn't a tablet - stop trying to make it one!

Game over. Do you want to play another game?

(...you know - material ui is the new shit. I wonder what are they smoking, i would _love_ to try it too)

This is one of my favorite parts of OS X, and it's sad that there aren't many Linux DEs doing it: the persistent menu bar. It's always expected that it be right at the top of the screen, and utilized. Firefox and Chrome, by default, have survived hamburgerization...because it can't be hidden. Sure, an application that doesn't use it at all (yuck) can just not implement anything beyond the application name menu, but something like a browser has no choice but for the bar to be visible.

I guess, we may be confident to see the menu bar dissappear in a not so distant future iteration of OS X. With a bit of bad luck and some rather cynical retroverse distortion, it may be replaced by a Hamburger (as in the Xerox Star Interface, which introduced it to indicate the anchor for the single menu in the entire system.)

Placing the cursor in text by imitating the new cursor catch and place game of iOS will be also fun. Probably so much fun that quite a number of people will stay longer in the office to play with it.

That said, the menu bar has already taken serious damage by the introduction of OS X, when the first menu item became the variable width application name. Before this, the File and Edit menus and their essential menu items were exactly on the same place in every application, rendering the most essential commands a matter of muscle memory, quite like a guesture.

Desktop software has never recovered from Office adopting the terrible Ribbon element instead of proper menus and toolbars.

Have you run into lots of desktop apps using a Ribbon style interface, other than Office? Microsoft didn't do themselves any favors by trying to patent the design and keep an iron fist on it.

I do see a lot of Windows software and utilities that have tried to copy it; there's a set of Telerik controls implementing most of the pattern.

It's just unfortunate, because it started the trend of reducing and hiding functionality. With less examples of what high density software looked and worked like, people don't even know what they are missing.

Then they briefly tried to push the one interface for every device holy grail with Windows 8 and UWP. Thankfully that was abortove, but it's still poisoned things more and hurt information density in its fallout.

How does the ribbon hide functionality compared to menus? I’m a big fan of it because it exposed functionality like styles, using large buttons that intuitively show how the function works, to people who were never going to independently discover them and customise their menu for easier access.

> How does the ribbon hide functionality compared to menus?

Did you tried to use it ? For every god damn thing i try to find in the ribbon i lose 5 minutes. That's what i call "hiding functionality" .

Yes, I’ve been using it almost every day for a decade, so I know where everything is by this point – but even when it was released I liked it almost immediately, because it surfaced useful style, paragraph spacing and table editing functions that used to be buried in menus. I prefer the ribbon interface in LibreOffice now, too. Next time you get lost (and it still happens to me just as it did in the menu era), I suggest using the search box instead of your memory of Office 2003.

Ribbons aren't easily scannable, because the controls don't line up. Menus line up vertically, toolbars line up horizontally. Granted multi-level menus are evil because what you're looking for may be in a submenu.

If Ribbons are implemented well, they're great. But often they're not, and then they're terrible. It takes a lot of work to properly organise a Ribbon for a complex program. Microsoft nailed it for Office, but programs like AutoCAD are still not well done.

Microsoft have put together an awesome guide to the Ribbon. Even if you're not making one, it's a great read that really goes into the reasons it exists: https://docs.microsoft.com/en-us/windows/win32/uxguide/cmd-r.... I came across this page by chance when I was in high school, and I think it was what sparked my interest in user interface and user experience design.

Foxit PDF reader does it and it's annoying as hell.

The Ribbon is great. At it's heart it's like a menu and toolbar, but with big recognizable icons, differently sized touch targets depending on importance, and always-on-screen familiarity.

It would work really well as a "menu" for a touch-based OS, even.

No, I cannot disagree with this more strongly.

There's nothing good about it that you couldn't do better with customizable toolbars before.

If there are ways to customize it now, I'm not smart enough to do it anymore.

> If there are ways to customize it now, I'm not smart enough to do it anymore.

Right click, customize the ribbon. You can set it up however you want.

Or just set it to show only tabs and treat it like a fancy menu bar.

Or hide it completely and just use the command search feature Office apps have in them now. It is like a command line with amazing auto-complete. Alt-q to jump to search.

Or just use the Quick Access Toolbar. I have it set up with all of my most used commands that I know the icon for.

I don't see the connection between the ribbon and the awful Metro design style. They couldn't be further apart in every way. It might look unfamiliar, but it still packs a ton of functionality into a small space, emphasizes the use of deeply structured alt menus and has tons of customizability. None of which shows up in Metro.

Also, I want to know what brain dead MS designer decided that multitasking different control panel windows was a Bad Thing, and put the whole settings section into a single window that can't be separated. Do these people even use their own OS?

I spend a huge amount of time making sure that my users get to experience fully-functioning UI on desktop, while not being overly limited on mobile. It's a HUGE part of my workflow, but honestly it's the only way. I can't stand visiting an 'app' website on my desktop only to find that it's 600px wide because it IS the app.

I was pretty blown away to see that Dropbox's hamburger menu opens to take over the entire screen, including on desktop. In other words, they only considered mobile in their design. You can try it even if you don't have an account, just go to dropbox.com and click the hamburger in the upper right.

How is that not pancakes? But ye, it's way to big. If they have to do it fullscreen, atleast put the menu options in the centre of the screen not downwards ...

I see this happening a lot. "Mobile first!" doesn't mean "Mobile only!"

I've had to conclude that they think (or hope) that everybody not using a phone is using a tablet.

Yeah, but surely the vast vast majority of dropbox's business has to come from PCs. Given how hard (and wrongly) they are pushing to be a collaborative space for work documents, I think I must be correct about this. I know it's cool and all to imagine everyone working on their phones, I just can't see that as a viable work process atm.

Their loudest majority is probably mobile then.

Not necessarily a majority.

but who uses the desktop web interface? don't dropbox users use the desktop app?

Sarcasm is hard. Unless I'm mistaken, you're imitating Blizzard and Diablo Immortal. If not then I think your comment comes across as extremely blind to a large segment of users.

Some people use it by dropping files up to share with people, and don't use the syncing capabilities. It's a reasonably common use case, I've noticed.

Some people also use it like a USB dongle, and drop files up manually, then download them on computers they don't have installation privileges on.

I use it to sync the pictures and videos from my phone, which is then synced to my desktop at home, and then deleted from dropbox and thus also the phone. Automatically. If my phone gets stolen (or data deleted by some overzealous border patrol goon) it doesn't matter as much as it would have otherwise.

I do, because all the people at my office using the desktop app complain about bad syncs, sudden filling up of the local hard drive and general slowness

I was forced to use web interface after app reinstall because Dropbox free account now supports only 3 linked devices (and I have 7).

It kinda does though, because modern software companies tend to shift focus away from projects before they get completely "done". The whole reason for "mobile first" is the tacit admission that whatever is not the primary focus will often get drastically less attention because there's already something that sorta "works".

There’s truth in what you’re saying, especially in the “things mostly suck, so expect that” world.

But in the ideal world you’re wrong... it’s fundamentally easier to start with a mobile design and make some specific interventions to make it work well on desktop than to do the opposite.

Whether anyone actually makes those interventions is, as you indicate, somewhat unlikely.

I personally think "mobile first" should mean we force the devs to use a mobile device as their work computer until they all realize how fundamentally crippling mobile UI is and we all agree to stop shoving it on everyone.

Looks like it only does this on sufficiently "small" screens. This is what it looks like on my external display: https://imgur.com/a/DJBiYJ6

Oh you're right. Interesting. My browser is 1134x1080. I guess maybe in today's world of super high res phones that could be considered a "mobile display".

Any CSS that doesn't take screen density into account is doing it wrong, though.

I don't really understand what's wrong with a full-screen modal menu, or how that means "this is for mobile." It's true that it looks a bit stark on their homepage where there are only 2 links, but it's probably a fairly sensible extensible design that works well when there are many more links or things to see.

It's fair to criticize specific visual and UX design details, but it's a bit much to take this as any indication that Dropbox doesn't care about desktop users (especially Dropbox) or that everything is being "mobilified."

Because I'm trying to read off some details on a document I have open on the side while interacting with the menu and then it goes full screen and obscures everything.

Fair enough. I only see two links in the menu, so surely the menu can quickly be viewed and closed.

> I don't really understand what's wrong with a full-screen modal menu, or how that means "this is for mobile." It's true that it looks a bit stark on their homepage where there are only 2 links, but it's probably a fairly sensible extensible design that works well when there are many more links or things to see.

You never used a computer? There is a thing called a mouse which you have to move to the bottom of the screen to select the bloody thing. I'm shocked that Alan Kay did had the glorious idea of the full screen menu.

> It's fair to criticize specific visual and UX design details, but it's a bit much to take this as any indication that Dropbox doesn't care about desktop users (especially Dropbox) or that everything is being "mobilified."

How to say it without resorting to medical condition: to throw away years of working things in GUI and to come with an "UX" which has the functionality of windows 1.0 shows some mental development problems to say the least.

I don't think it has anything to do with resolution. I'm getting the same thing as the parent on my 1366x768.

Edit: It's actually the width that matters. After reducing it a bit it will show the hamburger menu.

The media query for the hamburger button is max-width: 1366px, so you just met the cut off. I couldn't find a media query for the menu itself, but I suspect since the hamburger button is what allows it to get created, they didn't feel the need.

Mine's at 1280 and it's full menu, almost entirely whitespace. Makes it kinda fun though, feels like finding an easter egg.

Hamburgers and Heiroglyphs, GUI of the future!

At least Dropbox hasn't gone full in on the Heiroglyph side of the fad, they still have text labels for buttons, so you can understand what they do by looking at them and reading them.

The hieroglyphs are beyond me. It's literally making up your own alphabet not to mind language. Might be the ultimate sign of the over indulgence of developers/designers. Each gets their own perfect world which we must spend the time decrypting

I wish at least they used the egyptian ones. I 'd love to learn their language

Press (Stork) to save!

Then you'd love OS/Iris, which died a long time ago but was later resurrected!

https://en.wikipedia.org/wiki/Osiris_myth#Death_and_resurrec...

Maybe they try to save on i18n.

That is pretty easy - you don't need to support text in multiple languages, which is both difficult, a tad expensive, and slows down your velocity (I think the last on is the biggest issue).

Man that page is garbage.

With JavaScript disabled, you get a big white page of nothing. Not to mention the full screen menu as was mentioned.

They probably designed it to work on the CEO's tablet and called it a day.

I can't see the hamburger menu.

But here is the upper part of what I can see: https://imgur.com/eRFnDo6

Isn't is beautiful? How the image layer is put over half of the text line?

Oh, and the right part (white background) only appears after a couple of seconds. At first it shows the continuation of the blue background scree, with 2 'menu' items: "Se connecter" (log in) and "S'inscrire" (sign in). So if you try to click on "sign in", by the time your click is registered, it ends up on the white background item which appears at the same place: "Télécharger" (download).

Wonderful.

In fairness, Dropbox's primary interface on desktops is "my files are there" and nothing else. I can understand why they'd spend most of their UX effort concentrating on platforms where it's not as transparent.

you used to click on the icon in the taskbar and get a menu only. Now you get a full popup with a home screen , advertisements and "notifications" (even though windows has integrated notifications. Never used it

I have never clicked on this before, but mine on Windows 10 isn't what you described, and even if it was... why would I ever click on it? My files are backed up and show up on other machines. They could put animal porn in the UI and I don't think I'd notice.

I've just seen the new Microsoft Office. On the desktop. With a big screen. But just pressing the File menu makes the document you work on completely disappear behind something useless.

I've tried Word and Excel, both behaved so.

Insane!

> I've just seen the new Microsoft Office. On the desktop. With a big screen. But just pressing the File menu makes the document you work on completely disappear behind something useless.

That's been the case for I think a decade now. Maybe longer.

To be fair, now days they make really good use of that space. Earlier iterations of that UI were pretty bad, but a giant list of recent documents, a list of recent document templates, and all the main commands to do things (save, open) are positions to the left where they have always been.

Given that Word now days auto-saves (about time!), the main thing I'm going into the File menu for is to switch to a different file, at which point the file menu showing me a list of, well, files, is pretty useful to be honest.

Also it has a search box, and search boxes are nice, I welcome UI changes that allow for more search boxes in the world.

I think it auto saves only for Office 365 or SharePoint based documents. It definitely doesn't auto save my documents when they are local or on a network share.

One Drive files as well.

Strange it doesn't do local, but I imagine there are legacy reasons for that. Everything in Office is old. Often it is crazy well engineered, but decisions set in stone decades ago still linger on.

> Often it is crazy well engineered, but decisions set in stone decades ago still linger on.

YMMV

Well engineered and coded decisions made in the age of single user desktops are still well engineered. Hard for someone in 1995 to predict the need for a file format that supports real time coauthoring with automatic saving and complete revision control of every edit.

Esp. when the target machine the code had to run on was a 486 with a slow hard drive and a tiny bit of memory.

Of course Office updated their file format, but they probably did it a little bit too soon. I always wondered what would've happened if they had waited a few more years, released in 2007, so probably design started a few years prior. Talk about just being a smidgen too early for a world of unlimited(-ish) data and always on internet connections.

Oops.

> I've just seen the new Microsoft Office. On the desktop. With a big screen. But just pressing the File menu makes the document you work on completely disappear behind something useless.

Stupidity is contagious. And Microsoft's idea of innovation is copying what the others do and brake it beyond all repair.

> I've tried Word and Excel, both behaved so.

> Insane!

I'm shocked you used such a © "word". The new way to have fun is to be in a meeting and see the presenter trying for 5 minutes to find something in the ribbon menu.

> Stupidity is contagious. And Microsoft's idea of innovation is copying what the others do

Microsoft Office switched to the "backstage" menu in Office 2010, long before full-screen menus became a popular pattern. Office 2007 had the app menu, which is the same as the File menu in the Windows File Explorer. In my opinion, the app menu is far superior simply because it doesn't change the context and it's easily dismissable.

> The new way to have fun is to be in a meeting and see the presenter trying for 5 minutes to find something in the ribbon menu.

The Ribbon is far superior to the old teetering tower of random menus, toolbars, and panels. Everything is in a single logical place, with (almost) every button having a text label and many having intuitive icons. And now it has a search bar.

I'm fine with the ribbon, but good God do I hate the full screen save & open menus.

They've obscured the "save as" functionality and really push one drive saving. Even as a one-drive user, oftentimes I just want to save a file on my disk in a specific spot without backing it up to the cloud.

I really wish Libreoffice Calc was a better Excel replacement, because it feels like Excel gets more dumbed-down each year. I get a UTF-8 csv emailed daily from a vendor, and I still can't get Excel to treat it properly. It only applies that setting to that specific file, I can't change it system-wide. Maybe it's possible, I found it easier to just write a python script to convert the files to windows-1250 like it expects.

The ribbon is far superior to the old menus though, this week I needed to create a shudder PowerPoint presentation. Don't think I've needed to do that in 5 years, but it was easy enough to find all the functionality I needed. Plus, the search box actually shows where the buttons are located, so you actually learn how to use the program instead of relying on search. I remember the mess the Office 07 menus were, any time there was a need to use a feature that wasn't familiar was a chore. I remember actually needing to Google for some excel feature and needing to rely on screenshots of the menus to find what I needed. I don't get it, since Visual Studio has far more complexity buried in it than Word, but I've never had anywhere near the difficulty finding features than I did with the old Office menus.

There hasnt been a good excel released in 20 years. Libreoffice is pretty hideous, gnumeric is maybe a bit better. I rely on excel key shortcuts which didnt change even when the menus they referenced disappeared. Libreoffice has some but not all of them. Google sheets does better but my dev network is airgapped so that doesn't help.

What I refer to is the picture like here:

https://superuser.com/questions/642978/change-the-behavior-o...

Clicking on the File menu gets you

"Protect Document -- Protect Document"

"Check for Issues -- Inspect Document"

and

"Manage Versions -- Versions"

effectively as the only "obvious choices" with two different wordings (why?) and with a lot of "stuff" left and right "Properties" "Related Dates" "Related People" ? It's specifically not obvious that the "menu" is in the blue area on the left and that the three buttons are... what are they actually? How come these three came to be the major actions one would want to do in the File menu?

And why the document disappeared like I'm on the small mobile screen? I'm aware now they did that for many years, still... it's... confusing on the big screen.

> just go to dropbox.com and click the hamburger in the upper right.

Wow that is sloppy.

That's not true, their desktop breakpoint is 1024px. If the window is wider than that, there's no hamburger menu, and clicking "Sign in" reveals a menu that takes up 1/3 of the right side of the screen.

This is what I see: https://i.imgur.com/nmi0FHQ.png

For me the hamburger button's media query is set to a width of 1366px.

I don't have a dropbox account. I do notice no matter what I always end up at `https://www.dropbox.com/?landing=dbv2`, so I may somehow be being put into an AB test. Even an incognito window does that. The name `dbv2` might imply a new version of the site that is in the works? Shrug.

Ah, I think you'll see what I see at https://www.dropbox.com/?landing=dbv1

Yours looks like https://www.dropbox.com/business

I went to Dropbox to try this out but got redirected to "Dropbox Business" which doesn't have a hamburger.

And then spent nearly 5 minutes trying to get back to "Regular Dropbox".

But wait... I'm at the root of dropbox.com. Is... Dropbox "non-business" dead? I stopped using it a while ago because of their inexorable push into "enterprise" features that I didn't care about, but I never expected them to outright kill the product.

But our A/B testing shows a 5% increase in engagement by making it crappy.

But then we actually had a 13% decrease because our test was fundamentally flawed. Also, we don't really understand the statistics behind a/b testing ¯\_(ツ)_/¯

Oh yeah, i was pressing all your buttons randomly because none of them were obvious what they do. I was surely engaging with your app very much, while swearing.

Voice recognition is still a few revisions off, thank you for your patience.

Meanwhile it seems like very few people in the industry actually understand statistics well enough to not shoot themselves in the foot. The number of times I've seen accidental p-hacking is absurd. Having to explain to people why they can't stop running a trial as soon as the numbers look good is like pulling teeth.

https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3204791

Yeah, I always use desktop old reddit on my phone... I like small text I zoom in on.

What I hate even worse though is endless scroll... give me damn pagination with buttons! So I can reload the page and not worry about losing my spot, or accidentally scrolling away... let me use the back button on my browser!

oh yeah. i get that with twitter. view a video, close it and you ve lost your page . i dont know why this even happens but it s more annoying than i could ever imagine. My solution is to switch to chronological-view-only, so at least i can tell whether i ve seen my whole feed

i.reddit.com

This is fallout with the idea of "mobile-first" development... while I get that it caters to a larger demographic, it is blinded by its own crapulence in regards to _functional_ UI/UX.

> I hate that desktop designs are imitating the nonexistent information density of phones.

I think a lot of mobile designs underestimate the natural information density of phones too.

I think the biggest culprit is "tap targets", meaning that everything has to be bigger than your fingertip so that it's easy to select on a phone. It leads to crazy amounts of white space around everything. This ruined the presentation of tabular data on the internet, like stock financials or nba.com box scores. They're waaaaaaay harder to read than they were 10 years ago.

Google is to blame for this. If you run your site through their usability wizard, it will warn you when links or buttons are too close together, along with insinuating that you won't rank highly for mobile devices if you don't fix it.

The vast majority of the world use mobile devices as their only computing device, and the office workers and power users such as us are minority users and thus get minority attention. I think something like iPad OS getting laptop features and everyone targeting one or two mobile OS targets will be the future in 10 years.

It's like proliferation of electron apps, or how even the prosumer SLR / MILC camera market is dying as smartphone cameras get better and better.

> The vast majority of the world use mobile devices as their only computing device, and the office workers and power users such as us are minority users and thus get minority attention. I think something like iPad OS getting laptop features and everyone targeting one or two mobile OS targets will be the future in 10 years.

If you consider that a tablet is a "computing device" then my washing machine is also one.

> It's like proliferation of electron apps, or how even the prosumer SLR / MILC camera market is dying as smartphone cameras get better and better.

They get "better and better" because they are crap. As U2 said: even better than the real thing.

I really dislike menu bars. Seas of text with tiny hit targets strike me as not usable at all, and studies have shown this.

This! Somehow I feel that the desktop UX model of 90s were actually conceptually much better than what most modern apps are doing and it's a shame that instead of doing incremental improvements on that (sure, it's far from perfect) we've thrown out the baby with the bathwater.

The reason why scrollbars are not a first class UI widget in mobile UIs is because they are finger based, which means every manipulable widget needs to be finger sized. Making scroll bars finger sized takes up way too much screen space in mobile UI, so they were regulated to a passive progress bar indicator and the scrollbar actions were moved into swipe gestures instead. With that change you lose the ability to precisely skip around your location, probably because a desktop usage study showed that most people didn't use the the scroll bar much once scroll wheel mice were available.

Scroll bars only work with precision mice actions, and now desktop UX is legacy UX, mobile UX concepts enroach onto desktop UX, because developers do not like to create special cases, because that is more work and thus more $$$.

> The reason why scrollbars are not a first class UI widget in mobile UIs is because they are finger based, which means every manipulable widget needs to be finger sized

And yet, several mobile apps I use routinely have found a way to make it work reasonably well. They use a slim line of a scrollbar to indicate where you are, and if you touch the scrollbar then a larger finger-friendly handle appears which can be used like a traditional scrollbar.

But I'm much more forgiving of UI compromises that are forced by the limitations of mobile devices. It's when those same compromises are used on the desktop that I have a problem -- that's just degrading the UX for no good reason.

Testing on my iOS device, I don't think it's an app thing, but UIKit standard thing that comes with scrollviews. The bar doesn't become finger thick, but it's still usable. I've been using iOS devices for a long time, so it shows you how hidden it is.

I agree it's sloppy on desktops, but it helps to understand the reasons why the sloppyness happens. Kind of like electron apps :p

I believe it’s new in iOS 13

Way back in 2008, the Music app made use of this sort of feature (baked into UIKit). Table views can be "indexed," meaning you get A, B, C, etc down the side, and can swipe on it to move faster down the list.

Contacts still has this.

I've heard this from several sources, but never been able to make it work. What apps, and what version of iOS? Is it new in iOS 13?

It's pretty much everywhere you can see the scrollbar. Just press on the scrollbar for a second (so that it seems deliberate), then drag it around.

It is new in iOS 13, for the record.

Scrollbars as a passive visual indicator are still extremely useful though.

iOS 13 has actually made the scrollbar into a first-class UI widget. You have to long-press on the scroll thumb while it's visible, but doing so gives you direct manipulation over it.

Thanks for pointing this out. Don’t think I would have figured that out.

"Don’t think I would have figured that out."

This, unfortunately, describes half the features of iOS. Apple is deeply in love with the undiscoverable UI - and then NOT DOCUMENTING it.

Drives me absolutely crazy.

This one is actually in the iOS 13 release notes you see when selecting Learn More in the software update screen.

That said, I stumbled across it by accident first.

Samsung has implemented scroll bars in their browser and other Android apps and it works perfectly fine without being clunky or obtrusive.

Touch scroll bars don't need to be precise because most of the time we can just scroll normally. But it's nice to have an ability to jump to a specific part of the document in a consistent way.

I work on a desktop. Often a 40" display, in fullscreen. The thin scrollbar is really painfull to use.

Let's not dance around in a circle - Apple started this with their "minimalist design." First on desktop and later with mobile.

Fact of life is everyone copies Apple. Apple doesn't copy and paste things. They'll think deeply before copying. And will modify things to fit their unique needs and situation.

Unfortunately, those who copy Apple switch off their brains and pursue the "looks" ALONE, ignore the "feel" and just mess everything up.

The next big culprit is Google's Android with their let's just hide everything - hamburger.

Microsoft, the final boss, attempted to beat Apple in design. The result was Metro UI - a beautiful, text heavy, flat, design (they changed the name too many times to keep track off.)

Microsoft turned off their brains as well. The result:

They hid every thing - scroll bars, status bars and title bars. Wasted valuable screen real estate - empty spaces everywhere. Created a much hated start-screen with scroll bars that moved horizontally instead of vertically. And to make matters worse, moved in the opposite direction of scroll wheel. (Perfect if navigating with a finger but very confusing if using the mouse. It clashed with 20 years of mouse knowledge)

That's why people hated Windows 8.

Both Apple and Google riffed off Metro design. Google calls theirs material design, don't know what Apple named theirs.

If this article was titled, Who Killed Ports, 3.5mm audio jack, repairability, screws, or upgradable ram? The answer would still be Apple.

Summary: Prioritizing form over function is cancerous.

> Fact of life is everyone copies Apple.

And I really, really wish they'd stop it.

A small number of well-considered UI/UX design schemas really is a useful thing.

Apple is, generally, good in many particulars. It is extremely conservative, with only two desktop models (Classic Mac, 1984 - 2000, OS X, 2001 - 2019, and ongoing, which is to say, longer than the original interface). It is ultimately completely antithetical to how I work in desktop, and I find myself vastly prefering Linux and the option of selecting a desktop environment. Ironically: Windowmaker, based on the foundations of the OSX Aqua desktop, NextStep.

Increasingly it seems that Apple is stuck with implications of design decisions made decades ago which correspond increasinly poorly to current practices. The high cost of cycling between applications in particular. No, I don't want ALL FLIPPING WINDOWS of an app to pop up when I switch between the primary current working windows of two different apps. Or workspace flipping as I switch / open new app windows.

Apple's third OS UI, iOS, has a tremendous number of shortcomings. As do virtually all other mobile interfaces. I'm coming to the conclusion that any screen size < 10-15 CM is simply too small to be useful, and prefer a minmum 9-10" (20-25 cm) devices myself.

> A small number of well-considered UI/UX design schemas really is a useful thing.

Yes, in principle. But it becomes less than useful when the common schemas get in my way, which has been an increasing problem over the past few years. The aping of Apple's decisions appears to be a large part of why.

Agreed.

As jwz has noted, UI is different, in that changes to it provide relatively little payback:

[I]n the case of all other software, I believe strongly in "release early, release often". Hell, I damned near invented it. But I think history has proven that UI is different than software.

https://web.archive.org/web/20120511115213/https://www.jwz.o...

(HN-safe archive link.)

The bulk of present UI/UX metaphors can be traced directly to the Mother of All Demos, now over fifty years old:

https://www.invidio.us/watch?v=yJDv-zdhzMY

That said: from MoAD onward, the use-case has been of a desktop (or very-desktop-like laptop) system, with keyboard, pointer (mouse, tablet, light pen, etc.), used at something vaguely resembling a desk or table.

It's not even the screen size of mobile devices in raw resolution that's so different -- early GUI computers had as little as 640x480 (VGA) resolution or less. The original iPad was 1024x768.

But no keyboard, and finger-on-glass manipulation, call for a very different set of control and design considerations. A mouse pointer can have pixel-accuracy control. Fingers cover a probabalistic smear of screen space tens of pixels across and hide the screen elements they're affecting at the same time.

(And that's before you introduce such joys as screens repainting and repositioning as / before you complete a touch action.)

The real question is how to achieve concensus for new design and to void bad entrenched patterns across a highly fragmented design industry. Ranty flipping blog posts are one possible solution....

This is "easy" to solve!

Come up with a really good UI concept, implement it, and see everybody love it.

Avoid patenting it through the nose, and let everyone copy most of it. This is how it will become the standard and win the world adoption.

It seems like you'd need millions of R&D budget spent in a large part for public good, or at least to seriously change the market landscape without a guarantee you'll dominate it.

The design and usability history of OS X is actually a bit more diverse. The first iterations ("lickable" Aqua) had serious issues, favoring buyability and wow-effects over everyday usability. (E.g., as Doc icons increased in shape on pointer proximity, they also dodged the pointer as the intended drop target in default settings.) This consolidated with 10.3 and 10.4 was the first real usable version of OS X, leading to 10.6, Snow Leopard (2009), Apple's study in excellence. From there on things come a bit more in flux however, with increasing inconsistencies, until the Jony Ive era gradually introduced what may be considered a new interface. (If you compare the current UI to early iterations of OS X side by side, they are probably further apart than these earlier revisions from the Classic UI, which were at least sharing the same Hunan interface guidelines.)

Apple's interfaces have definitely evolved, but the changes have tended to be, compared to other platforms, minor and reserved. The overall motif of menu fixed to the display (rather than windows), three window buttons, hotkeys, etc., haven't changed, though colous, styling, and some features (multiple desktops, dark mode) have come and gone.

My preferred desktop, Windowmaker, is even more conservative, as it's seeking to emulate the 1980s NextStep desktop, though there have been a few changes, and I noticed many of them sharply.

But contrast Microsoft, which had fairly sharp changes from the Windows 1 days to present, or the gratuitous changes tossed out by the major Linux desktops (GNOME and KDE), and it's no contest. Mac has an incredibly stable GUI. I don't much care for it but it's stable.

Apple at least kept the minimal UI design where it belonged -- mobile. They also have a history of getting it right - the sole reason for the iPod and iPhone's success was it's beautifully simplistic design, especially on the software.

Microsoft in a panic, went overboard and started the trend that our 26" desktop monitors didn't need much on it. At least they corrected most of the issues by now, but many other companies haven't.

"That's why people hated Windows 8."

However, Windows 8 had an attempted solution for the specific problem called out in the article (the lack of a standard convention for getting an overview and jumping to a section of a long document or content scroll): semantic zoom. You could pinch-zoom out to get an overview of a view and then quickly jump to a specific section. (Scrollbars still existed but were only shown while using a mouse or pen. The mouse had access to the zoomed-out view as well, via a button on the scrollbar.) This was sadly removed in Windows 10 and I really miss it.

So, people may have hated it, but it solved the problem the article highlights, which today's "less hated" products don't.

I'm old enough that I started out building desktop apps with native widgets. You had consistent, easy to use UIs, and they were pretty much the same no matter what application you where in. Building the UI was considered they easy part. You could even draw the thing with drawing tools! Since the native widgets were designed by people who spend all day thinking about UI, user testing UI, etc everything pretty much just worked, and worked well.

Fast forward to today. There's no such thing as "native widgets" in a web app. So everyone rolls their own. Everyone keeps making the same mistakes over and over again. Everyone is chasing fads. There is no consistency from app to app. Today building the UI is the hard part.

(x-windows was the exception of course - you could choose from a dozen badly designed window managers and widget sets.)

> There's no such thing as "native widgets" in a web app.

HTML form elements are essentially "native" widgets, if only native to the browser. The problem with them is that older browsers lack support for many desirable widgets (e.g. calendar date entry), so some "UI-building" work is still needed.

Html form elements are laughably limited if you want to make an app - despite the years of using the web for apps, HTML+CSS is still only really usable for document-like pages.

This is not necessarily a bad thing, mind, there is something to be said about how huge browsers are becoming, and shipping a full widget toolkit would bloat that up even more.

There should be an HTTP-friendly GUI standard and we should have GUI browsers that implement all the common GUI widgets we know and love rather than emulate them with buggy clunky JavaScript/DOM/CSS. And be state-friendly to avoid having to use Ajax-like techniques to get desktop-like state.

The industry is simply missing a standard. Everyone was hoping HTML/JS/CSS would mature enough to solve it, but it didn't, despite being given 2+ decades. Developing in and using web standards for anything data-centric or highly interactive is a royal pain. You need rocket science to build a bicycle. Time for a dedicated bicycle standard.

There are any number of Web design failures which are traceable to either bad defaults or missing basic functionality. Standardisation on more of these elements, and maybe bisecting or further splitting current browser roles to (principally) textual content, commerce, multimedia, and applications, rather than cramming all four roles (poorly) into a single app, might help.

Good point! This is especially the case in iOS/Android, where almost every app uses their own completely custom UI and design on everything "just because". Remember when most apps used the iOS/Android OS-provided UI's? Customization was there when needed but dialog boxes, drop-downs, etc. were uniform across the OS.

This is why every app and its mother didn't need a UI/UX designer until the past few years.

>2. No flipping way to quickly navigate to the top or bottom of the post.

I don't know what the state of this is on Android, but iOS 13 lets you grab the scrollbar and go immediately to wherever you want in a document.

The scrollbar doesn't show up by default, so it's a 2-step operation. Scroll a bit to make the scroll indicator appear, then grab it and scroll faster.

Works well but I doubt many people are aware of it yet. People just assume you can't grab the scroll indicator in iOS since Apple took that feature away when they first launched the iPhone. Scroll bar has been non-interactable for the last 12 years of their touch devices.

>I don't know what the state of this is on Android, but iOS 13 lets you grab the scrollbar and go immediately to wherever you want in a document.

It's inconsistent on android, it works like that with most things but strangely not with some apps. You can't grab the scrollbar in Retroarch for example, even though you can see it. It makes scrolling through huge lists of ROMs extremely aggravating.

> It's inconsistent on android

Inconsistency in Android is one of its selling features if you are in to that.

I don't mind it most of the time. I wouldn't mind if scrollbars would let you utilize them consistently though or at least work like a scroll bar should all the time. To be fair though, I've used plenty of desktop software with inconsistent scrollbar behaviour to so...

> It’s inconsistent in Android

The only thing consistent about Android is its inconsistency, but even that can vary.

On a more serious note, I can’t find the date this was originally published (??? Seriously?!) so I don’t know what OS and browser versions were prevalent at the time of writing.

Top right corner over by the user info - 3 years ago.

I double checked by going to their other posts - it is the post-age, not the user-age, despite the "follow this user" button right next to it.

I did not know this. Another way to quickly get to the top of tables in iOS, assuming you're in an app, is to double tap the button of whatever page you're in in the navigation bar at the bottom. Some apps somehow disable this but it's the default behavior built into the OS.

You can pretty universally scroll to the top of anything on iOS by tapping the title bar along the top!

Works in Safari as well, though if the toolbar is collapsed your first tap will be to expand that, then the second one scrolls up.

Tapping the button of the tab you're on doesn't always scroll to the top; it does always scroll to the default position. In the "All Photos" tab of the Photos app, which is sorted with newest photos at the bottom, tapping the tab again will bring you down to the bottom, as that's the default position.

This is a special case; the default behavior is to scroll to the top.

On iOS, you can tap the top of the address bar to scroll to the top (I think since iOS 3.0 or so). This became 2 taps when they started hiding the address bar (one to show it, and another to invoke the "back to top" feature).

In Chrome (and probably other browsers), you can use CMD+Up to jump to the top and CMD+Down to jump to the bottom.

On various apps, with Google Chrome/Android being the worst, it can take minutes to scroll through pages / tabs.

I tend to have a lot of browser tabs open (well over 1,000 on desktop Firefox, which actually is fairly managable), and on Android there's simply no way to deal with more than ~20 (which I can easily hit in a minute or so).

Firefox/Android is far better in this regard, but on my device is so slow (literally minutes for UI responses) as to be unusable. Yes, despite numerous touted "performance" improvements. I'd very much like to switch. I simply cannot.

On Pocket, which is mentioned in the article and with a longer linked Reddit post ("It gets worse the more you use it"), it takes several minutes to scroll through my Tags list, with absolutely no search or keyboard accelleration possible. This on a tablet with an attached Bluetooth keyboard.

Tablet/mobile usability is complete crap, which is disappointing as tablet form-factor is otherwise promising. (Subject of yet another rant, this also on Ello.)

OP, FWIW.

> well over 1,000 on desktop Firefox, which actually is fairly managable How? :O What's your workflow?

Probably Tree Tabs, a vertical treed list of tabs on the side of the window.

Tree-style Tabs.

I tend to keep various projects/tasks organised by windows, and am increasingly using Incognito sessions not to hide anything in particular, but simply to keep myself from accruing yet more history.

I'm pursuing a wide-ranging set of research across numerous topics, and there's a large amount of material which seems at least potentially relevant. Cutting through what I've found and categorising it is a major part of what I'm trying to do, though how well I succeed is ... debatable.

There are also programming and sysadmin tasks, and other projects that I dive into. Those typically get some sort of state, whether it's a tab tree within a window, or a separate window.

Tabs are horrible. But they're better than all the other state-management methods which have come along from time to time.

On Android, the problem is that past a small handful of tabs, management, place-keeping, and showing relationships between tabs (parent/child, generally) becomes impossible. I've been trying to close out the open tabs I've got for ... a couple of years now. Sites have come and gone in that time.

Scroll performance in these scenarios is perfectly fine on an iPhone. I also tend to have hundreds of tabs up in Safari, not that I condone working that way...it's a sickness. Alas I've recently learned Safari's limit is 500. It's cumbersome but you can zip through them quite fast.

You seem to be describing a deficiency with your platform or your hardware, not with "mobile."

Exactly right.

> Lord knows, you can't actually grab the scrollbar and navigate with the bloody thing.

I read those words while actually grabbing the scrollbar and navigating with the bloody thing.

I actually prefer the current scollbar situation. I HATE when people hijack it but I prefer it to disappear when I don't need it. I can always scroll a tiny amount (magic mouse) to see my progress. I really do like the clearer look to UI's without the bar.

> No flipping clue how long the flipping post flipping is.

> No flipping way to quickly navigate to the top or bottom of the post.

I've literally never had either one of these issues with a disappearing scrollbar. OCCASIONALLY I have to try to grab it a second time but I'm more than happy with that state of affairs.

Also calling out Ello at the end is rather weak IMHO. If you really hate the lack of scrollbars so much then host your own blog...

This happens all the time for me.

When the page scrollbar is thin, and one of the page-width elements in the page also has a scrollbar on the right?

I am fighting on both mobile AND the desktop!

It’s extremely bad UI design when even on the desktop browser with a mouse and keyboard, I end up fighting to grab the page scrollbar.

And it’s not the page designers fault here.

So turn on scrollbars at the OS level?

That depends on the browser.

For Chrome, https://www.reddit.com/r/techsupport/comments/19glyg/how_to_...

A bad default that has to be edited by hand, in a tucked-away location.

This is awful.

> I prefer it to disappear when I don't need it.

My strong preference is the exact opposite of this. I want it to be there, full size, any time that scrolling is possible.

It seems that this is something that should be user configurable.

I can't help you on mobile but this is user configurable on macOS: https://imgur.com/kXDDjnF

I've only started using a mac recently and this drove me mad.

Thank you for making my life 5% better kind stranger!

You're very welcome! Since you are a new mac user you might also be interested in the list of Mac Apps I use [0]. Also if you don't use a magic mouse/apple mouse scroll can be jerky and annoying. For my gaming mouse I use Smooze [1] to help smooth out scrolling.

[0] https://joshstrange.com/my-mac-apps/

[1] https://smooze.co/

System Preferences > General > Show scroll bars: Always

OMG you just brought an enormous amount of happiness to my life! these disappearing scroll bars were a nemesis of mine and i have been blaming Chrome for this calamity. thank you so much :)

Which system/application is this for? Unless I'm missing something (which is entirely possible) it doesn't map to anything I use.

It's a MacOS system-wide setting.

Ah, thank you. I couldn't find it because I don't have any Apple machines.

For smaller screens, scrollbar-on-scroll (with a slow fade of 0.5 - 2s or so) would be a reasonable compromise, if the scrollbar were also a manipulation interface, and not merely informational.

The default Android PDF viewer operates in this manner, and provides a grabbable scroll control. Much of the rest of the app is otherwise less than impressive, but this would be a useful motif to pick up.

I'd offered the feedback to Ello in the piece as both relevant to the site specifically, and in the spirit of having long offered feedback on both good and bad elements of the site, at least for as long as there was any response to those (this stopped being the case around 2016, the site is now effectively dead).

I am a firm believer in the right to provide feedback on tools and services used. "Go self host" isn't a reasonable or scalable response generally.

> For smaller screens, scrollbar-on-scroll (with a slow fade of 0.5 - 2s or so) would be a reasonable compromise, if the scrollbar were also a manipulation interface, and not merely informational.

> The default Android PDF viewer operates in this manner, and provides a grabbable scroll control. Much of the rest of the app is otherwise less than impressive, but this would be a useful motif to pick up.

I wish I had an Android device laying around to test this on because I have a hard time believing any scroll-bar-manipulation-on-mobile would be anything but terribly finicky. That said I also don't find myself in a situation where I ever want to drag the scrollbar on mobile. I'm perfectly happy just using it as a progress bar.

> I am a firm believer in the right to provide feedback on tools and services used.

That was never in question.

> "Go self host" isn't a reasonable or scalable response generally.

Maybe "then host your own blog" wasn't the best way to phrase that, let me try again:

If you really hate the lack of scrollbars then there are a number of platforms (free, paid, hosted, self-hosted, the list goes on) you can move to that will give you the ability to turn on scrollbars.

Addressing the last: Ello's descent into UI/UX hell was among the reasons I abandoned it.

The Drive PDF Viewer (default PDF viewer on Android) scroll behaviour isn't perfect, but it's a reasonable compromise with minimal surprises (principle of least astonishment) for the platform.

I quite honestly prefer paginated PDFs in general for online reading, as they follow numerous predictable behaviours, with PocketBook and FBReader my preferred apps for this.

The Internet Archive's online BookReader app is also excellent, in its paginated mode. When set to portrait aspect ratios, it converts to a scrolling mode which I really dislike.

TFA is about scrollbars on mobile.

Principally, yes, but not exclusively.

...and hijacking scroll behavior. Please stop that and just use the defaults, like everyone expects. I'm tired of getting to a site and having it scroll pages at a time when it should just scroll a few lines. And then trying to get back to the correct spot is a pain. Who does this stuff and why do you think it's necessary? What problem do you think you are solving by doing this? I just hit the back button. I'm not going to struggle and fight the browser to digest whatever it is you thought was important enough to print, because you've made it unnecessarily difficult and complicated to consume.

I think this is the bigger issue compared to the design of the scrollbar. Personally I don’t really mind if the scrollbar disappears like it does on my Macbook. But when the website hijacks the scroll so I can’t use any gestures to go back or forward, that really pisses me off.

The usual culprit is some crappy Themeforest WordPress Theme the website is using.

NLM's remapping of vertical to horizontal scrolling is absolutely maddening.

When I indicate an up/down movement on a page, I expect the page to move top-to-bottom. And absolutely NOT left-to-right.

I instantly switch to the "classic" view, though disabling JS on the site has the added benefit of "breaking" the dysfunctional mode.

Any other remapping of scroll behaviour is also annoying as flipping heck.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4669363/

Maybe I'm a unique and strange person but my preferred scroll behaviour when reading a website is to middle click and move the mouse. Speed varies as the mouse moves so if you move just a bit you can get the page to auto-scroll at your perfect reading speed.

Of course, it's broken in a lot of sites, because many front end devs apparently don't know or care that this exists.

I'm born in the 80s so it's not totally objective, but I know there's something wrong when I boot up a win95 box and I don't miss 99% of so called innovation regarding UI.

I doubt it's nostalgia, even though i m old too. It's just a lot more efficent and faster to use, mouse and keyboard-wise. I often think to roll back to my windows 7 as a more sane desktop environment.

Some of the biggest offenses i 've seen are google's web apps. Total loss of hierarchy and replacement with elusive self-hiding buttons, even when it's clear from their use case that they should never have been hamburgerized. I 'm pretty sure, even in mobile, they are making the UX worse.

I'm mystified every time I try to find my contacts from Gmail. I understand that it's a separate app (I think), but did anyone stop and think that the two apps might warrant a bit tighter coupling (at least in menu placement)?

Google will use three different button schemes in a single app, nobody does product there.

KDE or plasma whatever its name, is basically unchanged from 1995 UI philosophy. The closest we can have today in 2019.

I run XFCE (via Xubuntu) on my home desktop for the same reason. The lightness is a bonus.

Going off on a tangent, but KDE is lighter than XFCE these days.

Or at least https://www.forbes.com/sites/jasonevangelho/2019/10/23/bold-... has been making the rounds.

Very interesting news both regarding KDE perf and XFCE team shrink. I hope they can get some workforce.

Thanks, I hadn't seen that. That's quite interesting. I'd ended up settling on XFCE aound the time of the Gnome 2 / Gnome 3 / Unity schism. At the time, XFCE had felt a bit snappier than KDE. I've lately been thinking about giving KDE a try again, so this is good to know.

Considering that Windows 95 powered many people's first exposure to computers, they clearly had to put a lot of thought into user experience. And they did a fantastic job.

I definitely believe almost everything since has been "innovation" for the sake of "progress", regardless of whether it was really better or not. Usually novel but worse.

Scroll bars and top menus! One factor was the removal from the market of anything but 16:9 displays which sucks as well. I want my workstation UI back and to return the fischer-price bullshit on every OS. Windows and gnome are the worst, every thing they do in the name of progress makes computers harder to use.

I miss the simplicity of Windows 2000/XP classic, been downhill from there. Ubuntu Mate is good but is suffering from the encroachment of gnome also.

Also took 15 years for dark themes to become allowed again due to loss of theming functionality.

I can see the issues w/ Windows, but what's your beef with GNOME? It's the only UX I know of that actually makes a touchscreen usable for non-trivial tasks. (And many laptops feature touchscreens by now. Often with something like a fully-rotating hinge, to enable touch-only use.)

I'm personally very allergic to gnome overall. I find it painful.

I use a keyboard and mouse at my desk. Not interested in touching the screen as it takes more effort and would need to be cleaned more often. I want easily discoverable menus, easily used scrollbars, and a dark theme. Real work has taken a backseat to play. I have an iPad/Phone for that and enjoy that at the sofa. Two use cases, not one.

The same people who killed folders "because tags". Don't get me wrong, I love tags, but why on earth can't we have BOTH???? I don't always remember what I tagged something with, and searching through 8 years of work to find something is useless.

Tags would be okay as a folders replacement if they, you know, actually replaced folders. That is: if a tag can in turn have tags, and those tags automatically apply to the tagged files, then it should be trivial to not only emulate a folder hierarchy, but even improve upon it by offering more flexibility.

It'd also help substantially if more tag-centric applications were up-front about which tags are already defined.

Agree, but I would add that another advantage of folders over tags is that folders are more discoverable.

LoCCS:

https://en.wikipedia.org/wiki/Library_of_Congress_classifica...

Seriously.

(Or one of a very few equivalents, possibly the Colon Classification.)

Tangentially, I find it amusing that the Library of Congress classifies the Bible as "BS".

Err isn’t this a problem remembering your file hierarchy as well? Search is the answer in both scenarios.

Well no, the point of a file hierarchy is that it should be well, a hierarchy.

In other words each level should sub-divide your files into further obvious (whatever that means to you) subcategories.

Sure, but that doesn’t explain how you magically remember where you stored the file. I suppose it does help with making you unable to store files in the root directory and forcing you to store it somewhere with some meaning. This seems like it would just resolve into a documents directory and you’re back where you started with a flat bag of files that you need to organize arbitrarily.

Also, you can use hierarchies with tags too, it’s just a matter of interface. Example: click “photos”, and see all the other tags that are associated with photos, and in a separate view all the files with all the previously clicked tags. You can drill down this way. Order top level tags by number of files using it. I have never seen this implemented but it seems like it would be fun. I also imagine “photos” itself would be a special file type metadata tag.

EDIT: posted before fleshing out idea, it is fleshed out now.

>I have never seen this implemented

Great argument for tags! Sure, it's never been done, but it's possible that someone will re-implement the inherent properties of filesystems with tags in the future, so let's take your filesystem away now!

Hey I am just describing interfaces, I don’t care about how it hits the disk. Filesystems are certainly nothing to get excited about; links are a dirty hack to get around a forced tree anyway.

Very few classification schemes are unambiguously DAGs.

We're used to them for use with physical objects, which can only occupy a single space. When the search interface is the representation (as with files or data stores), cross-referencing across categories is readily facilitated, and (in the extremely rare instance of well-designed ontologies), useful.

I've created several collections using extensive tagging schemas, and not infrequently run out of tags. I've also been looking at various ontologies suggested over the years, dating to Aristotle, Francis Bacon, Diderot, and more recent (often library or encyclopedia) schemas. Some are useful, none are perfect, all incorporate biases and anacronisms.

Another feature of folders (directories), or physical storage locations (for physical items), is that the classification / relation is created automatically by virtue of where an item or file is located. For tagging systems, you've got the additional overhead of specifying (or creating a rules-based tag-generating system) to apply tags to content. Otherwise you end up with a morass of "My Document0012811.docx", "untitled", "unclassified", "misc", etc., content. Cataloguing itself is a role, process, and step, and requires maintenance. Even within library science, the tendency has been to move toward self-describing works, full-text search, and imputed relationships.

For a classic instance of tabs done almost precisely wrong, take a look at Pocket. My two year old rant on it is still completely applicable.

https://old.reddit.com/r/dredmorbius/comments/5x2sfx/pocket_...

(OP)

I personally believe that the proliferation of infinite scrolling/pagination has a strong correlation with the disappearance of the scrollbar.

In a world before infinite scroll, the bottom of the page was indeed the last bit of content on the page. In today's world, scrolling to the bottom of a page typically loads more content.

In this new world, I can imagine designers and engineers deciding to do the simple thing which is to remove the scroll bar since it's no longer usually an accurate depiction of where the users position is in relation to all the content available.

Of course all of this solely applies to feeds and not standalone pieces of content like articles. In the latter cases I'm unsure beyond minimal aesthetics why the scroll bar would be removed.

Could I ask a related favor? Stop killing the Page Up / Page Down and cursor keys.

Or put another way, if you have a main content section that is scrollable (and the entire page isn't), then please make sure the focus is set to that scrollable div on page load.

I loaded the linked article and tried to scroll with Page Down, and nothing happened. I had to click inside the article body first, and only then could I scroll it with the keyboard.

I think a lot of sites and apps are built by people who do all their scrolling with the touchpad, and they don't know that there are other ways people like to scroll and it is the designer/developer's job to make sure they all work.

Pretty brave for posting on a site that breaks the back button.

And doesn't have a constant scroll bar

Ello is dead.

It was, for a while, promising, useful, and interesting. I'd posted content there while that was the case.

(OP)

For fuck's sake, if you need to be swearing (and I don't think it's necessary at all), just use the actual words.

We don't swear because it's necessary to convey the base information about something. We swear because it's useful to convey our feelings/emotional state/anger about it.

And of all things you could comment, not only is this not related to the topic of TFA at all, but it's also inconsequential, to the point of being some personal peculiarity... Who cares if he uses the "real" swear words or not?

Yes, it is off topic and I don't expect a full treatment of the pros and cons. However, it is certainly not the first time that people have commented on non content of a submission, e.g. bad color choice, hard to read, style, etc.

The thing is I felt the amount of swearing was distracting, and not using the actual word then made these swears stick out even more, distracting further from the content.

Maybe not relevant to you, but it's what I felt.

I often do.

I've been (occasionally) experimenting with slight bowderilisation. For better or worse, less swearing tends to go over slightly better. I can assure you the feeling was there in the original.

My general administrative announcement (NSFW langauge):

https://ello.co/dredmorbius/post/dlz9c2z6x-tvkd7rxtxwpw

https://mastodon.cloud/@dredmorbius/100521241625879766

Thing is I felt distracted by the amount of swearing, and the fake swears just exaggerated that. I don't care about the swearing in principle, it just felt misplaced in your article, in particular in this censored form.

Swearing helps convey frustration etc. and I guess that's what you were going for, but it doesn't actually work for me (as the reader) in all cases and in particular when it's "redacted".

Not that my opinion matters. I didn't intend this comment to be taken as serious discussion item and maybe I should have just not commented.

Fair points.

I wrote that some three years ago, among a series of rants (I don't recall the others that accompanied it), moved to extreme annoyance, and after looking at the initial result, decided to replace the initial language to tone it down a bit.

Today I tossed it into the HN submissions queue following an earlier item on similarly obnoxious and undiscoverable UI/UX, and was surprised to see the item take off. Planning which HN submissions will succeed is not a high-probability endeavour.

The best way to read the language is "yeah, the author was pretty annoyed when they wrote this, but decided to tone down the effect without removing all references to that annoyance entirely".

I've written highly invecitve-filled pieces elsewhere. For public or private consumption. It's when I stop any swearing that feedback (and usage) is quite likely to stop -- the outrage is a measure that I still care. Users past caring don't remain long.

I agree that scrollbars should be preserved. However, mobile screens are small and the classic scrollbars we were used to a decade ago take up precious space.

Actually, mobile scrollbars cause problems for me pretty often. My fat fingers sometimes trigger the scrollbar instead when I mean to tap an element on the right side of the screen. It's very infuriating when you suddenly jump to some random spot in a large list and have to scroll to find where you were and try again.

I would love to see I think the solution is to come up with a design that indicates an element is scrollable. Maybe by darkening it at the bottom? A scrollbar could be toggled when the user scrolls far enough, then hidden again when the user stops scrolling for a while. Idk, I'm not a designer.

is it just me or anyone else? As soon as i open a page on desktop i always look at the scrollbar first.

all the time - to assess how long the text is and decide whether to read it

Surely this is related to a concurrent thread

https://news.ycombinator.com/item?id=21353920

It's the flat design trend. It is unintuitive by design.

It is.

I'd posted a link to the piece on that discussion and decided to toss it into the submission queue on a lark. This got more attention than I'd expected.

https://news.ycombinator.com/item?id=21356495

Agreed that these mobile-esque designs are out of control but some remedies are:

- In iOS Safari you can tap the area above the address bar to jump back to the top of the web page.

- On iOS 13, you can finally press-hold the scroll blob and drag it in a sensible way.

- When there is a keyboard that lacks page-up/page-down/home/end, using the Command (Apple/Cloverleaf) key with up/down arrow tends to be equivalent to home/end.

- Fortunately a lot of web sites that have messed-up layouts still support Reader View, which can restore some sanity and standard controls.

> It's not so bad on desktops, and I can restore it there anyway via CSS.

I wouldn't recommend trying to modify DOM elements via CSS or extensions if you can avoid it. You're going to run into some broken web experiences.

Regarding scrollbars, you can set your OS to always show them. As a web dev, I'm out here trying to advocate to never screw with scrollbars.

Having no scrollbar could be descrbied as a broken web experience in my opinion.

And I actually think that begin able to modify webpage to your liking is important, even if it means breaking the page in the process (I'm kind of used to it with umatrix)

I do agree that in an ideal world, every webpage would be an amazing web experience, but unfortunately that's not the case, even with big websites with amazing UX desginer

Can you share a webpage that actually hides scrollbars? Aside from having a crazy amount of padding and shoving a scrollbar behind some div, how can you actually hide scrollbars cross browser?

It's literally one line in CSS to hide scrollbars.

You can also design the page to endlessly ask for new content as one scrolls, but never display a scrollbar...

In the popular browsers, how do we kill this misfeature?

OSX hiding scroll bars can be surprising for devs. You try your app on another OS and realize you’ve got several unintended scroll bars. I always turn them back on when working on OSX.

Yup, I learned this the hard way. We need to advocate for this within the companies we work for. Turn on scrollbars!

> I wouldn't recommend trying to modify DOM elements via CSS or extensions if you can avoid it. You're going to run into some broken web experiences.

If I'm resorting to extensions or CSS to modify DOM elements, it's likely the web experience is already broken from my point of view. It's not something I do just for the fun of it.

It's nothing to do with the web experience. Turn on your operating system's scrollbars.

The restyling is local only restyling, for my own use, and within my local browser stylesheets (generally using Stylus), only.

I've explicitly made that point in the example Codepen showing that styling:

https://codepen.io/dredmorbius/full/QwXBEO

More generally, I've long held that Web design as practiced isn't the solution, it's the problem.

https://old.reddit.com/r/dredmorbius/comments/256lxu/tabbed_...

I've been toying with the notion of an FYWD browser, standing for "Fine Young Western Dinosaurs" or "Flip Your Web Design", where some of the letters in "flip" were originally different.

Sometimes I scroll down articles quickly just to make the hidden scrollbar appears so that I can estimate the length. Very annoying.

However I do agree that removing scrollbars on Mobile has one advantage: I simply can't accidentally move to the bottom of the page, which (no idea why) I do it very often...

I couldn't agree more. Screw hamburger menus, scriptified scrollbars, and buttons that don't work properly. "Mobile first" is fine, but jeez, don't break my desktop. I mean, srsly, what do you think you're playing at?

You can grab the scrollbar in iOS 13. Not sure how recent that feature is because I don’t really miss being able to do it on mobile so I don’t bother trying. Which of course is not to say it’s a bad feature to have there.

> I swear heads have to roll. Literally.

Author had some good points, but that escalated quickly.

I think he meant heads will have to scroll.

I swear "eyes" will roll. There, fixed :)

They can remain attached to their bodies, if you prefer.

Ubuntu 16.04 Mate regions for resizing a window with te mouse is too small. Scroll bars are often too small. If I have to move my mouse back and forth trying not to miss...its not a good gui interface.

For me the worst example is Google Play Music on my phone. The scroll bar is there, but when you first open the app it's hidden underneath the pane at the top of the screen - you have to scroll down by dragging the window before you can even see it. Then if the window contents are long enough, as in when you're looking at the entire song list, you can't actually grab the scroll indicator unless you get really lucky - the sensitive part of the bar is much smaller than the visual indicator.

It flipping drives me mad.

Fast navigation = less engagement time = less ad revenue. Case solved.

I actually don't mind the lack of a scrollbar in the mobile version of Pocket. There are times where I don't necessarily want to see the progress of the article. The web version of Pocket seems fine to me.

What I really don't like is the Wall Street Journal mobile app which has an unreliable scrollbar, presumably due to ads. Articles frequently end well after you've scrolled to the "bottom" of the scrollbar so you can't ever tell how far through you are.

An old dumb phone of mine had a mini trackball I found immensely useful. The firmware (proprietary) was total garbage, but moving between menus was a breeze. Having a trackball rather than a touch interface means that icons and text don't need to be larger to be hit by a finger; all the user does is move the trackball until the relevant object is focused or text highlighted, then use it. Voila, lots of space saved for actual information.

The whole page appears as blank to me (Windows, Chrome, no addons, incognito mode).

I suspect that actually reading the article is unnecessary at this point.

It's a React SPA with no SSR. Why people do that for content only pages, I have no fucking idea.

Not to mention the scrollbar slider is almost invisible on windows 10 with the dark theme, and there are no way to fix it.

That's an awful detail.

There aren't even any scrollbars on Hacker News. There are no scrollbars anywhere.

Firefox may be to blame, but Ubuntu gets final responsibility. Firefox in Ubuntu 18 simply doesn't do scroll bars. In case it is an unexpected culprit, I switched to XFCE and then uninstalled a bunch of Unity stuff, but that shouldn't break Firefox.

Fix?

There are from where I'm sitting. I guess install Windows 10?

There's something very wrong with your install.

There are plenty of scrollbars here on xubuntu 18.04 and xubuntu 19.04

I don't know, maybe it's Unity aftereffects. I've been on xubuntu (xfce ubuntu flavor) since before 14.04 came out. It never suprised me, which is the killer feature. Also scrollbars.

I guess it's not that important as this article doesn't display a scrollbar either. /s

It mentions that at the bottom actually.

I see one.

A big part of the problem is it's not possible to customize the appearance of the scrollbar in most browsers. It's pretty much the only thing you can put in a webpage that can't be changed with css, so designers avoid them.

No scroll bars on iOS makes it look like features aren’t available and no indication that you need to scroll.

https://tyler.io/perfectly-cropped/

If you want to have a good scrolling experience with quick navigation for long articles on mobile try Opera browser.

Opera double tap zoom is also the only one that can relayout non-mobile friendly desktop paragraphs to mobile format.

This. Opera Android shows a small UI thingy (kind of hovering transparent scrollbar) on the right hand side (as soon as you start scrolling fast enough) which allows to super quickly scroll all the way down.

Thanks for the recommendation, though really this should be consistent for an entire OS or at least application framework (native and web), not something that varies app-by-app.

Though if Opera are getting this right, good on them.

I want to know why Android has absolutely ridiculous scrollbars that change size as you're scrolling. This is the stupidest thing and it's been this way forever. It makes them absolutely useless.

After the 5th "flipping" I stopped flipping reading the post.

Do I have the flippin' friend for you: https://news.ycombinator.com/item?id=21356928

Also see, from earlier today: https://news.ycombinator.com/item?id=21353920

Who needs scrollbars when I've got my trusty PgUp & PgDn

Lucky you. I'm noticing an increasingly shrinking ability for them to their intended job lately. Sigh

Many sites just don't bother to assign focus to the main content region. (I don't think they even realize they're doing it wrong.) Page up/down work if you first mouse-click there, assuming you can find a place to do that without the page hijacking it.

I'm glad scrollbars died on mobile. Right hand scroll bars are a PITA for left handed use on a touch device.

Apple has actually brought useable scrollbars to iOS devices for the first time in iOS 13.

Which would be funny as from what I recall, to answer the headline, Apple led the pack in removing scrollbars, from OSX back in about 2011.

I wonder if part of this (the web part anyway) comes from a desire to monetize content paragraph-by-paragraph. With simple HTML scrolling (which browsers are really good at) it's hard to report what content is actually being read. With JS-controlled pseudo-scrolling, that's easy but it makes for a lousy UX.

That tends to result in highly-paginated articles. In the worst cases, approaching paragraph-by-paragraph "slideshows" or pagination. An extreme anti-pattern.

Idiots removed the triangles at top & bottom that you could use to scroll without dragging. I think they did it because they are stupid and evil.

They did it because the buttons took up room, were used by nearly nobody, and (in the case of the "jump to top" / "jump to bottom" buttons) destroyed the user's scroll state in a non-undoable way that was too easy to trigger.

On touch devices, that's frequently the case.

Some form of multi-stage action (select/confirm) and a scroll-undo / scroll-backnav (return to prior location w/in doc/screen/page) might help.

Small-screen design is hard, and there are a number of distinct breakpoints at which different concepts do and don't work.

What OS?

Even without the arrow buttons I can scroll without dragging, just click the scrollbar wherever you want.

Other idiots also made the scrollbar thinner, harder to see, and hidden unless you purposefully move your mouse cursor to the side of the page. With a mouse and good eyesight, it's not that big of a deal, but I would imagine it's an accessibility issue looking for a place to happen.

ello breaks the back button on my browser

i have to click back 2x in order to go back to my previous page.

Iphone

From the story: "today's kids".

This is a pretty fluff piece. There isn't really much explored here.

The only interesting idea I can write out is that the author says that knowing the length/size of an article/book is intrinsically good. I do wonder if that is necessarily true.

Maybe one can argue that knowing the length of a written word has resulted in less people engaging?

This is absolutely terrifying thinking. You seem to believe that readers are passive consumers of a stream, and should not have agency in their consumption of said streams.

You got me. Readers "should not have agency in their consumption of said streams." Clearly that was the point of my question posing the alternative. You nailed it.

What I nailed is the idea of back-pressure. Removing the consumer's ability to use back-pressure mechanisms against excessive stream provision is an anti-pattern.

Providers who want to maximize "engagement" do not care.

This is known as "spam".

>This is a pretty fluff piece. There isn't really much explored here.

It's not supposed to explore anything, be amusing, or whatever.

It's just a rant about a certain specific UI trend.

>The only interesting idea I can write out is that the author says that knowing the length/size of an article/book is intrinsically good

That's not "the only interesting idea", in the sense that there were supposed to be more but are missing, or they are there but are not interesting.

It is literally the one and only idea the author wanted to convey with this post. They didn't set out to explore some vast expanse of UI design ideas. Just to write a rant about the lack of scrollbars/indication of position.

>I do wonder if that is necessarily true.

Yes it is.

>Maybe one can argue that knowing the length of a written word has resulted in less people engaging?

That's a problem for the content creator, not for the users...

For one, I remember a Google talk where they mentioned how the completion rate of e-books were higher than physical ones. One of the theories which stuck with me was that the reader not being aware of the end of the book could have been* a driver.

Yeah, this might be true, but that's a problem for the content provider I think -- or even not even for them (since you have already bought their book).

How did they determine this? Last I checked, physical books didn't have telemetry.

Yes, one can argue that, at which point one would be considered a marketroid and not trusted.

"Maybe one can argue that knowing the length of a written word has resulted in less people engaging?"

Boy, doesn't that just have the distinctive ring of an uncomfortable truth... at least for some of these cases.

I offer in counterpoint that engagement-focused sites have probably already determined that their content is "less engaging" if it's long enough for a scrollbar to appear at all. If it wasn't for the literal need to stuff 100%+ of the screen with ads, there'd probably be a lot of modern engagement-focused websites doing everything they can to fit on a single screen without scrolling at all.

When engagement is sold to advertisers, it's difficult to trust any changes to the UI.

The scrollbar should be sacred. Otherwise everything will wind up with sliding-window advertisements.

> Maybe one can argue that knowing the length of a written word has resulted in less people engaging?

Maybe, but valuing "engagement" over pleasing customers is a big part of what's wrong with a lot of software today.

"Lord knows, you can't actually grab the scrollbar and navigate with the bloody thing."

Do these people not have scroll wheels or scrolling gestures on their touch devices? I get it that some content is too long for scroll wheels, but it's just bad design if you are having to contend with data sets that are that large.

This just screams of old people screaming about change that they are not comfortable with... and I'm old AF.

I actually think he's making a big deal out of nothing on this one. If you have a site that gets long, it would be smart to create some kind of system for jumping between sections, or scrolling up a great deal. Like Google Docs does when editing a long document. I can jump between bookmarks and pages with the "knob" on the right. But most websites don't need anything like this. It should be the job of the developer of the site to realize their site DOES need this, and implement some kind of smart solution.

I’d argue scroll bars are often a “good enough” solution that you get for free and require very little maintenance.

In the past, navigation was the software's job, not the document's job to hack something in. You could expect navigation to be handled in a consistent way, and you didn't have to relearn something because you are reading a different site.

I know it sounds crazy, but what if there were an ubiquitous solution that everyone was already familiar with and had used many times in the past, a solution that required little thought from developers and behaved consistently everywhere?

Something like... well, a scroll bar, at least before devs started being cute and hiding them for the sole purpose of aesthetics.

That's crazy talk!

Because a scroll bar taking up real estate on the right side of a phone screen that is hard to touch because it's so thin is silly.

A knob on the right that's easily grabbed and barely covers anything is better.

Reply on news.ycombinator.com