Slashdot Mirror
Experts Say Gestural Interfaces Are a Step Backwards In Usability
smitty777 writes "Veteran usability experts Donald A. Norman and Jakob Nielsen wrote an interesting article lamenting the current state of the art in gesture interfaces. According to them, the lack of standards for interacting with these devices puts us on par with the '94 vintage in web design, when designers discovered they could make the buttons and UI look like anything they wanted."
puts us on par with the '94 vintage in web design, when designers discovered they could make the buttons and UI look like anything they wanted.
Hmm... this has given me some good ideas for an iOS app I'm farting around with. However, I can't find how to add faux-BLINK tagged text and Geocities-type spinning, flaming skulls in Interface Builder...
Trolling is a art,
What with all the OS companies trademarking the various gestures, there's no way they'll become standardized. Unfortunately.
It's not surprising that this has come about again. It has been roughly one full generation of developers since 1994. During that time, those developers who actually learned proper usability techniques either retired or moved on to other endeavors. They knowledge they acquired and the methods they developed have basically been lost to the sands of time.
Today, we have a whole new generation of developers creating this shitty software. They'll spend the next 10 to 15 years learning what the previous generation had learned. There'll be a few years of good UI design before these developers move on, at which time the cycle will repeat.
Gesture based interfaces are a bit of a mixed bag, if they are done well (see the iOS pinch gestures) they work very well, but if badly implemented you end up accidentally triggering them all the time. Despite the age of the classic "object" based UI designs, they are still the best control method (in most cases), just because you can see what you are doing by what you hit.
I'm not being sarcastic here, but this is why i've felt that the atrix i own is an inferior phone to the n900. In the n900, the upper corner always took you to the multi task screen where you could close the application out, and if you closed the app, it always worked. This was because it had a not-as-friendly-to-touch interface that was based of of linux guidelines. There was consistency, but if the button wasn't visible, all applications still responded to it (unless they were frozen, then a freeze popup would happen, allowing you to close).
This has been bugging me for the past few months with the android, and now i know why it just doesn't feel up to snuff. The android phone is the first phone i've ever owned that had mystery behavior.
It's not as if there aren't human interface guidelines in place for some of the gesture based environments in question. While not capital "S" standards, who cares?
Akin to the early days of GUI interfaces, we didn't have standards in the early days when the Amiga, Apple, Atari, BeOS, NeXT, Windows, OS/2, GEOS, CDE, roamed the earth. There was the Apple HIG, IBM's User Interface Guidelines and so on.
It didn't stop the awful Win95 interface from coming into existence, but so what. Let people create and innovate. We're all the better for it.
...the slashdot April fools ohmigodponies interface. It was the pinnacle of web design and nothing has come close since.
These posts express my own personal views, not those of my employer
I've got a single gesture in mind for folks who think that gesture-based interfaces are where it's at...
Actually, I do like the intuitive "pinch, spin, slide" type gestures with iOS, but for PC-based stuff, I can't stand a lot of the new, shiny crap folks are pushing. Removing useful things like status bars, and replacing intuitive "I don't know what I'm looking for, but I'll know it when I see it" menus with those "trying to view the Grand Canyon through a toilet paper tube" restrictiveness of these ribbons and such... it just really gets annoying.
The Digital Sorceress
seeing as it has a standardized set of gestures including a standard gesture area. But it seems to focus exclusively on iOS and Android...
"Waste not one watt!" - CZ
I thought with the title, it would be a social study about how gesturing at a computer like an ape instead of sitting down and calmly telling your equipment what to do (via text or speech) is a major steps backwards for humanity. How can people not realize that every new technology will go through a phase where everyone implements their own idea before the industry settles on a few good ideas?
Eggs
Milk
Bread
Cat Litter
Soda
Gesture based interfaces might not be perfect but they've sure expanded the use of smartphones well beyond the typical Treo and Blackberry crowd. The interface can't be that bad.
Now you kids with your loud music and your Dan Fogelberg, your Zima, hula hoops and gesture interfaces, don't you see? People today have attention spans that can only be measured in nanoseconds.
Look where all this talking got us, baby.
Gestural interfaces are ok on a touch screen, but when using a mouse, I find they're just inconvenient.
worldmobilenet.com -- World Prepaid Wireless Internet plans
"on par with the '94 vintage in web design, when designers discovered they could make the buttons and UI look like anything they wanted."
Yep, we've come a long way since then.
The best part about navigating Netflix and Hulu on my XBox 360 is the voice control--dead simple and unambiguous with straightforward voice command options on the screen if you need them. "XBox pause!" "XBox rewind!" You can mutter voice commands, talk slow (not too slow), talk fast, talk like a cartoon character and it just works. It feels like how the 21st century should feel like.
The hand gestures, on the other hand, are an exercise in frustration. The Kinnect is good at tracking your hand, but you still have to use a fine degree of motor control to move that little hand to the pause area of the screen. The requirement to use fine motor control for something that can be accomplished using a voice command or by simply pressing a button on a remote control is a STEP BACKWARDS. The gesture commands are OK for scrubbing through the videos, but only because you would be using a similar motion if you were using a controller or a mouse.
Bill Clinton: Pimp we can believe in. - The Shirt!!!
Gestural interfaces are ok on a touch screen, but when using a mouse, I find they're just inconvenient.
Yeah, mouse gestures were popular in Opera in, what, 2002? I think the difference is with a touchscreen they're the best you've got but with a mouse you have more expressive options.
My God, it's Full of Source!
OUTSIDE_IP=$(dig +short my.ip @outsideip.net)
Actually, with interfaces, everyone doing it the same way is better than a few of them doing it differently, even if those few are actually doing it in a way that is otherwise easier to use, because the point of an interface is for people to understand how to use it.
I'm not anywhere near the caliber of UI expertise as Norman or Nielsen. But there's a big advantage to pioneering a new physical interface: you don't need the language part of your brain. My 1 year old twin nephews can interact with their iPads with only the most basic of demonstrations of how a new app works. They can't read or write but they can follow demos of fingers creating action pretty well.
Is bringing along the old interface of mice & menus helping or hurting? I particularly like the new "swipe up" gesture to scroll down of a touchscreen rather than the traditional "elevator window" model of scroll bars where clicking up scrolls up.
They are absolutely to be commended for chastising developers that there is no easy way to discover actions if they are not intuitive; I'd rather they come up with ways to address this than just fall back on menus though. For example, Apple included an interactive tutorial for using the custom gestures built-in to Pages, Numbers and Keynote because they aren't discoverable at all. Some I've forgotten because I don't use them (and I'd have to re-watch the tutorials again to re-program my brain). But the ones I have picked up on are absolutely ingrained and effortless now. Unfortunately, built-in tutorials are the exception rather than the rule, and even when they are included they more trouble to refer to than a drop down menu. But there are ways to improve without eliminating gestures.
I wouldn't want to use the gesture interface when I'm programming during the day, but when I'm swiping through my early morning junk mail, RSS feeds, and to-do items, my brain feels far more engaged on my iPad than my desktop. It's almost like the touch gestures are autonomic and leave my (limited) higher brain functions alone to read though the fog (at least until my caffeine kicks in.)
I agree that people need to improve gesture interfaces which are in their infancy, but I don't think it's justified to throw the baby out with the bath water just because of long traditions.
Lets draw some parallels here... I'd say that point n click UI is most like that of the layout for a controller on a classic console. If you follow a standard UI you have certain buttons and menus that users can identify with. For example," _, [ ] and X " sit in the upper left or right hand corner of most application windows. Users expect these buttons(minimize, maximize and close) and use them regularly. Likewise, a classic controller layout like those from Sony and Microsoft includes directional buttons, face buttons and triggers. Game developers use the similarities in control layout to map their action buttons. Multi-platform games have near exact mappings and games within a particular genre use configurations that are similar to one another.
Contrast this to button mappings of a game on the PC platform. Developers have 108 keys and a mouse at their disposal. They can create and mandate some very confusing control layouts. Gesture controlled UI design has just as many, if not more, possibilities. As some users have mentioned, patenting gestures does not help create standards. This just means develops have to think of new ways(read: different buttons) for users to interact with their application.
Where's the point in all this? PC games can have some very confusing control sets. However They havent failed yet. Many gamers prefer them over consoles with a more limited set of controls. I think the confusion over gesture UI will fade and with time more people will learn to accept the nuances
That's like saying vintage '94 web design was a step back from menu and keystroke driven application design.
At least, that's the Microsoft theory.
On the other hand, everyone from Linux to Android to iOS disagrees, and consumers have a choice because of it(I.e. if you don't like the Windows UI, you can use any number of Linux UIs or the OSX one. Phone UIs give even more choice).
*Get your experts to evaluate the competitions apps/platforms. "Oh noes! Three mouse buttons! One mouse button! This simply will not do!"
Have gnu, will travel.
What I saw: Expert's Gay Sexual Interfaces Are a Step Backwards in Usability
Seriously? LoL
How much practice is required before you're considered an expert at these homo-erotic interfaces?
Is there skill quantization "tool", or perhaps a "Queer Eye" review?
Are the controller's or receptacles aesthetically pleasing?
Do lesbians with optional strap-ons have an advantage over the rest of us?
Are Expert heterosexual interfaces not equally as ridiculous?
I laughed for a good minute before I was disappointed by a second read of the headline...
Laptop touchpads are a prime example. I'm pretty use to the gestures on my Asus EEE. But my wife's Dell is very different. She can't use mine. I can barely use hers (i'm more willing to figure it out). And of course not all of the gestures work that well in the Linux. How do you even discover the gestures? They don't print a list of them in any easy to access place. So we use USB plugin mouse whenever possible.
Stupidity is its own reward.
In the article they say:
In comments to Nielsen's article about our iPad usability studies, some critics claimed that it is reasonable to experiment with radically new interaction techniques when given a new platform. We agree. But the place for such experimentation is in the lab. After all, most new ideas fail, and the more radically they depart from previous best practices, the more likely they are to fail. Sometimes, a radical idea turns out to be a brilliant radical breakthrough. Those designs should indeed ship, but note that radical breakthroughs are extremely rare in any discipline. Most progress is made through sustained, small incremental steps. Bold explorations should remain inside the company and university research laboratories and not be inflicted on any customers until those recruited to participate in user research have validated the approach.
I appreciate that they're important contributors to UI design, but their attitude is unrealistic to companies that are trying to ship products, make profit and gain market share. Companies spending too much time perfecting their UI design will go out of business while their competitors are shipping flawed but ultimately usable products.
Not so much really.
All mouse/keyboard UIs are essentially the same. *NIX and Windows both have it pretty standard that there is a send to taskbar button, an embiggen/shrink button and a close button in the top right. OSX does something weird which I don't quite understand, but the buttons are still in the top. There's a file menu at the top on all of them. There's a place on the bottom that holds identifiers for running or commonly accessed programs. Windows are square, and have predictable functionality across platforms. The left mouse button activates whatever you click on, right clicking brings up a menu.
All touch based interfaces are essentially the same too. Big selectable icon that launches the app. Notifications and at a glance info is displayed at the top. Swiping side to side switches screens to the side. Swiping up and down makes the list move. Long pressing on app icons moves the app shortcut around. Pinch to zoom, double tap to zoom, two finger twist to rotate, etc. All of that stuff is standard across UIs.
Sure, the animations and style are different, but the UI portion of it is more than style. It's the interfacing that's the important bit, not the fluff that makes it pretty. Imagine moving from iOS to Android and not only is the fluffy stuff different, but now you have to tap the bottom of the screen and then the top to scroll up, and you scroll faster by repeatedly tapping the top. Or long press to zoom, and double tap to exit the application.
There's a reason that anything that requires user interfacing follows whatever the big player does to some extent. It's called usability.
Keep on knockin'
https://robbiecrash.me
you should RTFA, they make some very good points and a very cogent argument.
The Cloud - because you don't care if your apps and data are up in the air.
Your summary is incorrect. It should be "...since they don't apply the basic principles of usability, which have nothing to do with particular interface metaphors or technologies, they are making simpleminded mistakes."
It has nothing to do with abandoning the old desktop environment, and everything to do with giving poor feedback, providing arbitrary interfaces that don't provide cues to the user, not providing consistent behavior in similar situations, etc. These are problems that have more to do with understanding how users think and learn and applying that to the interface design.
well... indeed. Interfaces should be self-evident. If we have to break out the manual to use them, we might as well go back to CLI.
The Cloud - because you don't care if your apps and data are up in the air.
Especially people in Ubuntu disagree. They really try all different changes to the UI, some of them seem to be there just for the sake of "making it different than the standard" and they fight tooth and nail against users who want to bring standard behavior back.
I wonder if the changes are backed by any kind of usability studies, or do they make them as they go along, because some of them are so retarded really no sane user or programmer would suggest them.
45 5F E1 04 22 CA 29 C4 93 3F 95 05 2B 79 2A B2
... at least it'll get the fatties up off their chairs and MAKE them do a bit more aerobic exercise, so it's not all bad...
As a retired UI designer, the referenced article reads like sour grapes. The reasons for the "trouble" as posted (lack of established guidelines / misguided insistence by companies / developer community's ignorance) sounds like whining from a standards organization. Old farts always sound this way when someone younger and more bold come up with something new.
I think therefore I can't be ~TTNH
Some years ago I was blocked in traffic by an idiot and when he eventually moved, I mouthed "thank you" at him. He proceeded to follow me home, knocked on my door, and screamed at me that I had called him a "wanker". Well, I might have thought it...a universal gesture for "thank you" is surely needed.
From scarped cliff or quarried stone she cries "A thousand types are gone, I care for nothing, no not one."
Any interface I cannot grep through and trivially script is a step backwards.
You don't understand! Actual experts have proclaimed that point-and-grunt...er, pardon me, point-and-click is far more empowering than any form of communication using language! After all, chimpanzees can point and grunt, so clearly its a superior interface for humans as well. Sorry, I meant point and click again.
You're not going to argue with actual experts, are you? I mean, are you some kind of Luddite who wants us to return to the primitive ways of our ancestors? (Or is it the UI experts and Human Interface Guidelines that want that? I get a bit confused sometimes.)
Funny - he doesn't like the decades old windowed desktop environment either. I have to be honest here, if you fault Donald Normans logic on something, you better have good reasons - he is a /very/ clear thinker indeed. Read "The Design of Everyday Things".
I'm a minority race. Save your vitriol for white people.
More like 'basing a UI paradigm on easter eggs may not be the least confusing way to do things'.
The lack of visual cues, lack of consistency (when are what gestures available or not?), combined with the ease of random interference with the gestures make them less than optimal.
Personally I think it works fine as long as there is a half-dozen easily distinguishable gestures. But imagine trying to use a gesture UI that incorporates most shortcuts, contexts and menus available when using more complex applications or desktops, which is where such systems are going.
In fact, it is opposite to Microsoft theory
Microsoft theory would say: "Since we have the biggest share of market, let's make the UI of our system different so it is harder for the user to swith to another solution"
That said, I do not think it is time to worry (yet). Gestures is a new input method and people are experimenting with it, see what works and does not, etc... once it is more mature I hope we will see a trend towards standarization.
Why can't
I'm guessing you didn't bother to read the article, which points out that there are certain ideas for interfaces (such as keeping them consistent) that make them easy to use, regardless of whether it's mouse & keyboard or touch.
Some of their "problems" are not only risible, but their proposed alternatives are downright ludicrous.
For instance, their model for "Discoverability" (sic) on a touch-driven interface is to add buttons and menus. That's right, it is not user experimentation of touching and interacting directly with screen objects, but the flat, indirect execution of commands via a textual proxy.
Some of their complaints are valid: it seems every app developer ignores even platform conventions and does his own thing, or misappropriates Desktop UI interface conventions (e.g., WTF are "+" and "-" buttons doing on a map application when the user has been trained to expect pinch-and-zoom?) But these are due mostly to individual developers. Apple has some fairly good guidelines on UI interface designs for the iPad and iPhone, starting with the recommendation to design it specifically for the device, and not port it from a different platform; that new and inexperienced developers is not their fault.
This will all be sorted out eventually, as the technologies mature and the interaction paradigms are cemented into the cultural consciousness.
The article authors' idea of advancing human-computer interfaces for touch screens is to ignore the potential of the technology and stick to what works for a desktop and a mouse. Sure, they acknowledge the need to experiment with new techniques--just as long as they keep within the boundaries of what is known and what works already, mainly the desktop GUI and mouse.
-dZ.
Carol vs. Ghost
They also make some very bone-headed recommendations based on some intellectually lazy assumptions; e.g., that command menus are the best and most appropriate way to allow for discoverability on any platform or form factor.
Carol vs. Ghost
Usability alone should not dictate an interface, but it should also be intuitive. Unfortunately, the very nature of gestures is unintuitive. There are no visual queues as to what any particular gesture will do, whereas a button is very clear. Click here, and 'this' will result. Gestures are just too vague. They are handy as shortcuts but a standard must be applied for them to be useful. The first versions of the new Google app is a good example. It was all over the place, where a lot of it made no sense and seemed to add no value. Why must I pull down on the screen to see my 'apps' when a simple button that labeled apps is much more effective and intuitive? Another example was searching. When you are searching you probably don't need an apps button. They eventually incorporated both so that you actually had an apps button and various menu items disappeared where they were unlikely to be needed. That meets both usability and intuitive needs. I wasn't clawing at the screen trying to remember which direction to slide the screen to see my apps, and I now recover that space when I'm searching with only search related items on the screen.
Gestures by themselves with no visual queues are pretty useless to a laymen or someone unfamiliar with the interface.
Although I am a former accessibility consultant, I wasn't in it long enough to really have an expert opinion on gestures and whether or not standards would help anything.
What I can tell you is that guys like Nielsen make their money by auditing everything under the sun against these standards. In my day it was 508 and W3C (which are still valid today, but more easily satisfied, I think).
Some of what we did was really useful stuff in terms of educating other developers about how people would access the web with screen readers and some really fairly easy techniques to accommodate them and other people with disabilities.
But a lot of what we did was 'take this federal grant money to audit this state or local community college's web site so that they won't be sued under the ADA.' Your tax dollars paying for us to go around to institutions that could barely afford web developers (as they operate on your tax dollars) and tell them that yes, their web site could definitely allow them to be sued under the ADA.
It's definitely a dilemma though. There's a pretty easy argument that says that educational institutions should make their entire application process accessible. But as the technical guy in the process, I found that most of what went on was non-technical people talking to other non-technical people about how great the standards were and patting themselves on the back. We did a lot of 'sorry, you're not accessible' but very little 'let us give you a seminar on how to be accessible.'
I never met Nielsen, but he would make the news in our corner of the world a lot by going after some company or institution, declaring them inaccessible, and then hope to make a wagon of cash by getting hired by said company. That's how the industry is set up, though, so really I can't criticize him for it -- he would (rightly) say that too many people were totally ignorant of accessibility and so would use his reputation to put the spotlight on it and sometimes profit along the way. I just wonder how much stuff like Section 508 actually advanced usability.
And as anecdotal evidence goes, my two-year old was able to figure out how to use the photo viewer on my phone about as quickly as an adult was, when they were each introduced to it for the first time.
NOTE: Those gestures are very different than the trackpad gestures for "reply all" and their ilk, which are completely artificial.
You're special forces then? That's great! I just love your olympics!
You didn't RTFA, did you? Here are the useability problems TFA discusses in a nutshell:
Non-existing signifiers
Misleading signifiers
Feedback
Consistency and Standards
Discoverability
Scalability
Reliability
Lack of undo
It's a good article that goes into detail, and anyone designing any interface for any device should read it.
Free Martian Whores!
The bigest problem I have with MS products is the lack of consistency, and I don't just mean not following standards. Why can't they keep menu items in the same menu in an upgraded app? Buy a new version of a program you're familiar with and you ALWAYS have as big a learning curve as if you'd bought a competing product. They don't follow anyone's standards, not even their own.
Free Martian Whores!
"Discoverability" was in the Webster's dictionary since 1913. And I agree with him, plain text is always the easiest. I've always hated unlabeled icons.
Free Martian Whores!
The article does have a few points. But I find their problem with the menu button and the long click on Android to be a bit of a stretch. "How is someone supposed to know what is there?" Come on, the right click of the mouse is the same. Sometimes the items in the right click pop-up menu change depending on what you clicked on, or what application you are in, or if you are viewing the list of emal, or have one opened.
-- ssoorrrryy,, dduupplleexx sswwiittcchh oonn.. -Quote found on actual fortune cookie.
Yeah, because that's so intuitive. I mean, all physical objects in the real world work like that! It's an obvious metaphor!
P.S. I can't find a moron here to double tap.
P.P.S. People who run the subject & body together are imbeciles who ought to be boiled slowly.
Confucius say, "Find worm in apple - bad. Find half a worm - worse."
Already wrong, sorry. My linux is set up to have none of those things.
Yes, you've customized your installation so that it doesn't fit the standard. What is your point? That your personalized, non-standard UI proves that there is no UI standard?
Keep on knockin'
https://robbiecrash.me
(1) Tooltips - no such thing as a "mouseover". No (built-in) way to show them.
(2) Scrollable regions (frames, divs with overflow:auto) - if you don't show scroll bars, the user has no way to know that there's more content in the region that they can scroll to. Not so bad if the edge of the region visibly truncates something. Then you know that there's some more content there. But what if the edge of the region occurs between 2 paragraphs?
I'm not saying that these problems can't be solved. I have my own (site-unique) mechanism for tooltips and am in the process of coming up with some (site-unique) CSS to clearly highlight scrollable regions. I'm just agreeing with the authors that it's a big step back not to have any automatic feedback at all about the existence of extra content.
Yes. There is no UI standard. You seem to be confusing "standard" with "most common way people are doing things at the moment." It's like when I tell people that I'm not going to open their Word attachment and that, if they want me to read their document, they should send it to me in a standard format. They complain that Word is a "standard" and that Windows is a "standard" and what am I talking about.
So what defines a standard then?
Keep on knockin'
https://robbiecrash.me
There are several mechanisms. The standard for the format of an email message is an "rfc". There is the ISO. There are standards such as that for PDF files, published by Adobe. If there is a standard for a GUI, I'd be interested to learn about it.
So, if Microsoft were to put out a standard that says a GUI has all of the things that I stated, would that appease you?
Keep on knockin'
https://robbiecrash.me
That doesn't make any sense. How could a company establish, unilaterally, a standard for "a GUI"?
It doesn't and they couldn't, neither could a standards body.
So a standard-UI would have to be a de-facto standard that came about due to usage patterns and widespread industry design principals. So, while not a codified Standard with an RFC number, the UI I described is a standard UI. Standard as defined by the English language definition, that it is the 'normal' thing based upon popular usage.
Your pedanticism over the word 'standard' is the issue here.
Keep on knockin'
https://robbiecrash.me
A standards body can't establish standards? You've lost me.
I'll partially concede your point about pedanticism, though, and that you're using "standard" in a common, colloquial sense that would pass without remark in other contexts. It's just that in this context (Slashdot, etc.) the term has other associations and I thought it was relevant to point out that there are no standards, in the more "pendantic" sense, for GUIs (unless there are). This is more than just a pendantic distinction. The fact that there is a standard for PDF but not for DOC (for example) is important to understand. If we could get the computer-naive to appreciate the issue, the computer ecology would be better for all of us.
I just mean that a standards body couldn't create a globally definitive Standard for a GUI. It's impossible, because there are too many platform and context specific things that do not cross over unilaterally.
There are no officially defined standards I'm aware of, and I looked when we started this discussion, for GUIs across platforms. As this discussion points out, there are a bunch of platform specific standards and best practices, but nothing universal.
I agree that everyone that is significantly involved with computers should know that defined standards are critical to the entire computing ecology. But aside from us, nobody really needs to. They don't need to know how the stuff works any more than I need to know about the inner workings of accounting or marketing or any of that other crap. It's our job to understand this stuff and make it work, it's the accountant's job to make sure I get paid, and marketing's job to make sure I spend that money.
Keep on knockin'
https://robbiecrash.me