GUIs Get a Makeover 540
jcatcw writes "From Xerox PARC to Apple to Microsoft, the GUI has been evolving over the years, and the increased complexity of current systems means it will continue to change. For example, Microsoft is switching from dropdown menus to contextual ribbons. Mobile computing creates new demands for efficient presentation while the desktop GUI doesn't scale to larger screens. Dual-mode user interfaces may show up first on PDA phones but then migrate to laptops and desktops. Which of today's innovations will become tomorrow's gaffs?"
Touch screen talking pie menus (Score:5, Interesting)
I've been developing touch screen talking pie menus [piemenu.com] on handheld devices, like the Pocket PC. Pie menus work very well with touch screens, but of course the way they track and display and give feedback has to be adapted to the quirks of small touch screens. Talking pie menus give you audio feedback with a speech synthesizer, so they don't require a lot of visual attention and hand-eye coordination.
Talking pie menus make it possible to use an application without looking at the screen! That's important for mobile applications like GPS navigation systems, which people use while driving (despite all the warnings again it).
-Don
The Human Computer Interface (Score:3, Interesting)
I think a big problem is the mouse. The mouse is so great for so much, yet it falls short. I know they have mice that have practically a whole keyboard on them. I'd like to see that idea extended beyond the window manager also.
One thing that has really excited me recently is the Optimus dynamic keyboard [artlebedev.com] over at artlebedev.com. Thinking more about adapting the interface around the user and the software is important. A lot of that will be workflow analysis, such as "User A always saves before printing, so if they save, make the print icon easier to find and click." will be necessary.
A lot of what needs to be done the computer can do for us. The hidden options in MS Word are a good example of this. Although it was a support nightmare when it first came out, it really helps speed up the work when you are doing common repetitive tasks. This could be expanded to allow different hidden options depending on what you're working on. For instance, if you're writing a letter, addresses and envelope stuff should magically appear, but it should not show up if you're writing a scientific paper.
One thing that the MS monoculture has brought us is a somewhat standard UI experience for most users. That would be impossible with 100 competing OS's. The web does not offer that opportunity except maybe through some toolkits like Swing (which sucks), or Ruby on rails with the prototype.js. The monoculture has stifled innovation, however, so I hope in the future there will be more people thinking about design when they make their interface and MS being open enough with this Aero stuff to allow designers freedom to make something new. I seriously doubt that will happen, however.
Re:I dont agree (Score:2, Interesting)
Re:It probably won't change much more (Score:2, Interesting)
I saw the multi-touch display wall at this year's SIGGRAPH. Playing with it is, obviously, worth more than looking at pictures, but you really have to watch the multi-touch interaction demo real.
How will it improve my productivity? (Score:1, Interesting)
The one thing I always ask when presented with a GUI change is, "How will it improve my productivity?"
Now, from looking at those screenshots, I have my doubts. That new taskbar idea looks like shit. Notice that the labels on most of the items are truncated after about six to eight characters. It makes it difficult to know what they're saying. I mean, look near the bottom. What site is "www.pl"? What is "www.ku"? Were I not a previous user of KDE, I would have no idea that "Konver" was the truncated form of "Konversation". Look at all the wasted space above and below the truncated labels. Compared to the old panel, that sidebar is a piece of shit.
What's that thing in the middle supposed to do? Can I click on those icons on the left to accomplish something? Why are there icons that look to be from Mac OS X, and appear to represent web browsers and email clients, doing next to a description of the Kopete instant messenging software?
And why so many textual descriptions about the applications? Being a long-time KDE user, I already know all that information. I don't want to read some bullshit marketing blurb about Kopete, I just want to use Kopete.
As for the shot of Konqueror, I don't see the need for all that gradient nonsense. I wastes a lot of screen space. What is wrong with highlighting the item selected on the left, and then only having a thin vertical separator bar like it currently is? The current look of Konqueror is very compact, makes very good use of the screen space, is comprehensible, and works very well. I don't see any benefit from changing it.
Like I said earlier, I ask myself, "How will it improve my productivity?" In this case, I don't like the answer I see. I like that KDE is trying to innovate. But they're going in the wrong direction. They need to focus on making better use of screen real estate. We don't all have 24" LCD monitors yet. On a 15" or even 17" screen, any wasted space is unacceptable. And the changes they do make have to have a purpose, hopefully increasing productivity. But it doesn't look like that's the case, as most of these seem to do the opposite. They add confusion and reduce clarity. That's not a good thing to be doing.
Re:The Human Computer Interface (Score:5, Interesting)
I don't know much about the present speech systems in OS X, but the older one in classic Mac OS had a "speakable items" folder that was mostly filled with AppleScripts. Speaking the name of any item in that folder would launch that item; if it was an AppleScript, it would do various thing. The system shipped with a number of useful scripts already built in: one of them was called "What time is it?", and all it did was speak (via TTS aka MacInTalk): "It's [current time]", e.g. "It's five oh four pee em." (Then again, I don't find this very useful because I've got a menubar clock, as all Macs have by default for ages, so it's quicker just to glance up there).
There was one really impressive script in that that would tell a number of interactive knock-knock jokes, called "Tell me a joke". So you'd say "Tell me a joke", and it would speak (via TTS) "Knock knock". A response of "Who's there?" would prompt it to select from a number of responses, and it would then listen for "[previous response] who?" after which it would deliver the appropriate punchline.
I just looked, and there is a Speakable Items folder and it has all this same functionality still. Runs a lot faster than it used to, too. Sweet.
Re:I dont agree (Score:2, Interesting)
Dhumb! (Score:3, Interesting)
The Apple example, handwriting recognition on the Newton, is a good gaff. Which is to say it isn't something that any rational person would look out and say "That's dumb. Don't do that." It isn't Clippy. It isn't Bob. It's trying to get the computer to adapt to the person rather than getting the person to adapt to the computer. The big win for Palm was that Grafitti forced the user to adapt to the computer. Our handwriting is the way it is (hopefully) so that other people can read it to. Typewriting is not a natural thing, even though some of use geeks reach WPM speeds that make it seem like it is.
When we're talking about verbal user interface gaffs, we'll find similarly goofy things, and we'll find things that made sense intellectually but didn't work in reality. That's what we call research, kids.
Re:Touch screen talking pie menus (Score:2, Interesting)
Re:The Human Computer Interface (Score:3, Interesting)
I was just writing about this above. Actually, voice-recognition has progressed considerably in the last few years, due to handhelds. Cellphone voice recognition is practically standard, now days. There's a few problems with bridging the gap over to desktop computers (less with laptops, though), the main one being that most people don't have a mic built into their system. Companies have TRIED with mics built into monitors, but that hasn't seemed to fly, except in the Mac world.
These kinds of things scare me. People become faster with computers as they learn repetative operations. Even if something is a little more confusing than you would expect it to be, people become quick at it because they know how to do it. Placing the printer icon in a different place after a certain operation may speed up operation, in theory, but it leaves the user constantly guessing as to where the options are going to be placed next. I'm all for customization, but let ME do the customizing, through actually doing the customizing, I then learn exactly where things are. Traditionally, AIs have always been very bad at trying to figure out as to what users want to do, and usually make the operation much more difficult as a result. Take arrow "Snap To", for an example. The thing is supposed to figure out where I want to click next, in a dialog box. At the same time, every time it auto snaps, I'm left going "what the hell just happened?" and searching for the arrow. which takes a lot longer than physically moving over to the dialog box and clicking the button. If anything, I think computers try too much to figure out what you REALLY want to do, and most of the time it's either disorienting, or just feels patronizing.
Re:The problem with guis is they don't work (Score:4, Interesting)
Change is bad (Score:4, Interesting)
I firmly believe that when it comes to GUIs, change is almost always for the worse. One reason for this is that once a set of GUI conventions has become established, change is disconcerting--you now have to accustom yourself to the new "look" or to the new way that the GUI works. That inconvenience is rarely repaid by the alleged advantages of the change.
As an example, consider the difference between the Windows 2000 and XP desk tops. Just how is the XP desktop better than the older one? I sure couldn't see any advantage to it. Yet, if you were to use the darn thing (and not switch to the "classic" view), you'd have to figure out again how to do a bunch of stuff you already knew how to do before the interface changed. This is progress? Even at the detail level, the changes are silly and unhelpful. Look at those three-dimensional window title bars. Why is that bulgy look better than the less obtrusive flat title bar of the old Win 2K interface? What convenience or information is added by the 3D bulge? Or how about the XP icon for video options--it's a screen with a flat paintbrush on it instead of the 2K screen with a round paintbrush and ruler in front of it. The two look different enough that it takes me a couple of extra seconds to find that icon in the Control Panel whenever I'm forced to use the default XP interface. It's not that the new icon is better or worse than the old one--but why ever change a familiar, easy to recognize icon? It's done to create the illusion of progress, of course.
Making icons look "cooler" in successive iterations of software is one of my particular pet peeves. Whenever someone releases a new version of their software, they think that people won't believe they got their money's worth if the GUI looks the same--so they jazz up the icons. Usually, this means adding more detail, even though this violates the basic principle of the icon: that it should be simple and easy to recognize. In other words...icons should be iconic.
That brings me to another reason why software publishers change GUIs. From the article:
Excuse me, but if you've got "exploded" features, then you do not have a problem that can be solved by a revamped GUI--you have bloatware. Clean up the mess, and start over.
I haven't seen these new "ribbons" MS is talking about for LongVista, but even the name is dumb. Look, the people at Xerox Park gave us the foundation of a great GUI, and there's no reason to change that basic set of visual metaphors until there's a fundamental change in the mechanics of the computer/human interface. The requirements for a good GUI are well-understood: it should be as simple as possible, it should be consistent between applications, it should use easily recognized familiar symbols and conventions. It most definitely should not change from one moment to the next according to the notions of some guy in Redmond who thinks he can anticipate what I want to do.
Re:It probably won't change much more (Score:1, Interesting)
Re:Touch screen talking pie menus (Score:5, Interesting)
Of course I've heard of Steve Mann's work, and his Gnu/Linux Wristwatch Video Phone [linuxjournal.com], which used pie menus (but didn't talk as far as I know). He built his prototype pie menu watch in 1998, about 10 years after we (Jack Callahan, Don Hopkins, Ben Shneiderman, Mark Weiser) published a paper [donhopkins.com] about pie menus at ACM CHI'88. But in 1988 (and 1998), not many people had hardware they could carry around that was suitible for implementing talking pie menus.
Speech synthesis requires a lot of memory to store a good voice, and speech enabled applications require a lot of task-specific scripting control (so they don't start talking and talking at length about something the user is no longer interested in). I'm using the Lua scripting language on the Pocket PC, to develop flexible speech enabled touch screen pie menu based interfaces, which will run on commonly available Pocket PC phones. (I've done a lot of Palm programming in the past, but that's a dead platform.)
Here's a video that Dave Winer [scripting.com] took of me demonstrating an example application: a remote control for "Rock and Roll [podcatch.com]".
-Don
Re:I dont agree (Score:5, Interesting)
My product an image manipulation system, has had contextual, ribbon-based selection of tools since 1990. We use a chapter/verse metaphor (click on one level of the toolbar to select the chapter, such as filters or geometric tramsforms, the next level slides into view which contains individual tools such as sharpness and feature removal, or ripples and rotations.)
This layout, like MS's "new contextual ribbon" puts what you need in front of you, and buries everything else until you need it. Our chapters function exactly like MS's "tabs" and our verses function as accessors for sets of tools -- basically, there are three levels to the GUI. We don't put the third level in the toolbar, because there are far too many controls for some tools (as many as 70 sliders, buttons, drop-downs) and it is (we think) a poor decision to always take large amounts of vertical space in an image-processing application. Dialogs let you move all that tool-consuming real-estate around. They aren't modal, though, so you can keep working.
This really is a better and more evolved way to work, and I commend MS on finally getting the point (although I note with some humor that they certainly didn't invent this methodology.) Of course I'm partial to it, having been building and using such an interface for well over a decade now.
The thing that seems to stick in user's craws isn't the difficulty (or "increase in complexity, as you put it) of such a layout, because there isn't any, really... but simply that it is "different." Change is a force for user discomfort, especially UI change. I'm not saying that UI's can't get more complex, they certainly can, but contextual ribbons are a simplifying factor, count on it.
We don't need any steenkin' new paradigms... (Score:5, Interesting)
If software developers just spent an extra hour to watch an untrained user play with their software... and their managers gave them a couple of extra weeks to incorporate what they learned by watching... that would have more effect on software usability than the introduction of new techniques.
The problem today is that so much software leaves you gasping with amazement at the seeming perversity of their design. It's been observed since the day Windows 95 was introduced that it is stupid to turn off your computer from a button labelled "Start." Microsoft has had over a decade and one, two, three, four, five major software releases to do something about it, and they haven't. If they don't get it yet, all the pie menus and gestures and voice recognition isn't going to help them.
You may cry foul because this isn't strictly speaking, a software problem, but will you take a gander at the button layout on this portable DVD player? [dpbsmith.com] In case you don't get it--it's so mind-boggling it took me a while to get it--the northeast button moves you east, the southeast button moves you south, and so forth. That's why every button has a little printed arrow next to it.
An awful lot of modern software design seems to me to be be putting little printed arrows next to utterly misplaced buttons.
Re:I dont agree (Score:2, Interesting)
-uso.
Re:I dont agree (Score:5, Interesting)
Then, if you think about the code for cells...in "evolved" eukaryotes, there are not only long sequences of DNA inserted from viruses ages ago, there are copies of genes that just don't work because they're mutated. Talk about junk code. But those sequences are dutifully preserved inside your very cells. It's a nightmare that even Microsoft would hate to dream.
Re:The problem with guis is they don't work (Score:2, Interesting)
A GUI is intended to make it easy and intuitive to find out how to do something when you are not sure how to do it (discovery made easy). This is its greatest asset and its greatest liability, because this encourages users to use the mouse and look for buttons to do things. Which is counter productive and slow, but makes it easy for a complete novice to find out how to do something.
The idea GUI makes it easy for the newbie to do something and as they use it more would teach them quicker ways to do things. For example, it would allow a user to find functionality quickly when it needs to be discovered, but then teach the user more efficient ways of doing that task. It could display recently discovered functionality, for example, 'cut' as a GUI button, but teach the user that they can use ctrl-C to achieve the same task. once the user starts to use the shortcut it could hide the GUI button, say after they have used the shortcut 5 times, saving valuable screen real-estate and not encouraging the user to always use the button. The user could easily discover it again if they needed, and perhaps lock certain buttons for complex tasks or things that they prefer to use the GUI for.
The unfortunate truth is though, that this type of functionality would only sell to 5% of the population (basically slashdotites), because the rest of the world thinks that something is easy to use only because it is pretty and has lots of nice buttons with easy to understand graphics (which is true for a system that makes discovery easy). They don't know what is good for them!
Re:Every thing has its time... (Score:2, Interesting)
But perhaps they should put a few hours into fixing the hyphenation algorithm. That's just laughably bad. happe-ning, all-ows, communica-te, vario-us, prog-ram. Every hyphenation is wrong.
Re:GTK+/GNOME file chooser disaster. (Score:5, Interesting)
Personally, I don't mind that interface. Besides, if that's your only problem with GNOME, then we must have it pretty good! I "strongly dislike" KDE's browsing system (one arrow left, one arrow right, one arrow up, one arrow is a crazy swirl, all so close together and so similar in appearance that it really gets frustrating at times.) And why the default is set to open folders with one click is beyond me. I have one program (Noteedit) that uses the KDE interface, and because of that, I didn't bother downloading all of the customization crap, so I'm stuck with it (if someone has a solution, tell me please!). Also, the taskbar/menu at the bottom always looks too cluttered to me. And the clock is just ugly. And why do they stack the window list in two rows? I came over from the Windows world, and was introduced to GNOME and KDE at the same time (I was playing around with SUSE and Fedora). I liked both the same and eventually my final decision came down to the GUI. KDE just hurt my eyes to use. It's a little hard to explain. All of the icons were so...BIG, and pixilated. And despite the fact that KDE looked a lot like XP's UI, I went to GNOME.
From what I can tell, people are about evenly divided on this issue. It's just whatever appeals to you. No, GNOME is not paradise incarnate, but to me, it's better. Besides, I sure you can customize that path chooser
But isn't that the beauty of FOSS? The fact that you can actually choose? Sort of like democracy, it's all the arguments that actually let you know it's working.
Re:I dont agree (Score:5, Interesting)
Back on the subject of software, the more the complexity is packaged into simpler modules, the more the system above it can be simplified. The end goal is to have modules of a stable complexity (like TCP/IP) forming together to create a simpler OS. The problems occur when there's a monolithic structure that exposes lower-level complexity at a higher level.
The future is Adobe (ex-Macromedia) Flash (Score:2, Interesting)
But neither win32, mfc, qt, gtk, kde, wxWindows can find the promise of separate the OS rigidity from the UI
Just Squeak [squeak.org] give us more freedom, but from a business perspective Adobe [adobe.com] is playing a very strong card with Flex [adobe.com]
Imagine a creative designer with all the freedom to create the best UI without more limitations!
We can see some real examples of Windows interfaces in flash like ScreenTime [screentime.com]
Re:GUI? Bah! (Score:3, Interesting)
It's treated me well so far. I find it a lot easier to deal with than a Windows-style taskbar; I tried to duplicate it when I had to work in Windows for a job a while back (I even bought a virtual desktop manager [astonshell.com]), but Windows's support for virtual desktops still seems sort of broken, so it didn't work as well there.
Re:Too much "innovation", too little scope for cha (Score:4, Interesting)
It makes sense when you understand a) the purpose of the "Start Menu" and b) the history behind it.
The Start Menu is the "one stop shop" for initial tasks in Windows - it's the UI element you go to (or are supposed to) for launching programs, configuring the machine, searching, help, etc, etc. It is (roughly) equivalent to Classic MacOS's Apple Menu, the NeXT Dock, and similar "do it from here" elements in other GUIs. Logically, in Windows, the "Shut Down" command belongs in this UI element and nowhere else (with the possible exception of a dedicated button on the taskbar, like Ubuntu does - although back in the day the problem then would have been wha icon to put on the button).
*Originally* (in the first "Chicago" betas), the Start Menu wasn't actually called the "Start Menu" and didn't have "Start" on it - it was just a button with the Windows logo, much like the GNOME and KDE versions. However, during their usability testing, Microsoft found that users couldn't actually figure out what to do when the system first booted and all they had was an empty desktop and taskbar, with a little Windows logo at one end and a clock at the other (I can't even remember if the clock was there at that stage). So the button got a label - "Start" - to signify that it was the UI element where you "started" to do everything.
First impressions count a lot, so if you take away the Start button most people will feel a bit lost and will have a negative experience. Thus people won't want to let go of Start even if it is in their longer term interests to learn something better.
It's interesting to note that in Vista, the "Start" label is gone. Presumably Microsoft's usability studies have concluded that the "Start Menu" UI element is now so entrenched, users no longer need to be taught what it is.
Re:Every thing has its time... (Score:3, Interesting)
I say this as a KDE enthusiast who has a background with being in love with GNOME: How come no mockups like the ones you linked and the ones found on many many other places online have not been adopted yet into KDE4? Matter of fact, why is it that KDE4 and QT4 itself chugging along at such a slow pace? I guess I should be grateful that KDE3 is still seeing so much attention to detail, because it's only recently been able to woo me away from my GNOME desktop.
Also, I have to respectfully disagree with the usefulness and attractiveness of those mockups. The first I can't even figure out. This is an overcomplication of the desktop idea, in my opinion. A taskbar doesn't need to display any of that information, and that task basket seems like a solution in search of a problem. For the second one, well, I dunno, I've been with a setup similar to that for a long time on my Linux boxes, except for the blueness... Call me crazy but I like and always have enjoyed how Konqueror looks in the first place, and I have very rarely seen a mockup that improves upon its interface. KDE-Look has a lot to browse through. A lot of mockups I see around are basically trying to make Konqueror look like Xandros File Manager [wikipedia.org], which in turn looks a lot like Explorer.
But yeah, like I was saying: There has been no actual work yet that makes KDE4 any different visually than KDE3, or if there is, I sure as hell havn't found any screenshots.
Re:Johnny Cochran? Is that YOU!?? (Score:3, Interesting)
I often wonder why people continually make this assumption. There is no evidence for this point of view, at least none that I've seen.
Sounds like wishful thinking to me.