A New Paradigm For Web Browsing

dsaci points out a New York Times article about how surfing the web may change to a more graphics-based endeavor. With the advent of devices like the Wii and the iPhone, the capability to directly control objects on a screen is becoming a popular and affordable technology. That, combined with immersive interfaces such as Piclens, could be the future of web browsing. Quoting: "'I've wondered for a long time why the computer interface hasn't changed from 20 years ago,' said Austin Shoemaker, a former Apple Computer software engineer and now chief technology officer of Cooliris. 'People should think of a computer interface less as a tool and more as a extension of themselves or as extension of their mind.' Voice, too, is finally beginning to play a significant role as an interface tool in a new generation of consumer-oriented wireless handsets. Many technologists now believe that hunting and pecking on the tiny keyboards of cellphones and P.D.A.'s will quickly give way to voice commands that will return map, text and other data displayed visually on small screens."

Comments

Yeah good luck with that

[Gothmolly]

Dragon on a reasonably powerful PC might work, but until you can nail 110% correct recognition, in a crowded area, in a shitty little mic on a 400 MHz ARM processor, don't bother. You don't want to start arguing with your PocketPC about traffic and directions: No, I said Springfield, not Slingblade! *crash*

The keyboard works, 100% of the time. Its easily understood. Its robust. It fails gracefully - you immediately see if you've made a mistake before submitting a command.

Re:Yeah good luck with that

[jay-za]

You are underestimating the practice that years of running Windows has given to the average user. That *crash* will not come as a surprise. There could be a market for a technology that turns the windscreen blue just before the actual crash. Finally BSOD will have a more ... real ... meaning.

Re:Yeah good luck with that

[mh1997]

The keyboard works, 100% of the time.

Tyess, thek eypboard isg thew perfeddct ddevicwe4 requirening litttttttttttle skil and is foo l profo.

Re:Yeah good luck with that

[calebt3]

Giving the computer bad instructions != the computer misunderstanding those instrutions.

Re:Yeah good luck with that

[ShieldW0lf]

Aside from that, for someone who is competent it's faster to type than it is to speak, and it's much faster to twitch your hand half an inch than it is to wave it around touching the screen.

Everything outlined in the article is leading away from integrating technology into your core capacities. It's about taking a tool and turning into a third party agent that you need to interact with as though it were some sort of person.

Making a more efficient computer interface means making the muscle movements involved more subtle, not replacing what efficiencies we have with new paradigms that require gross muscle movements and voice strain.

Integrating mouse gestures into the operating system and and moving to one-handed chording keyboards as a standard would be the right direction.

If the brainless masses want Fisher-Price toys, fine. But lets not pretend that Fisher-Price make better tools.

Re:Yeah good luck with that

[mh1997]

"Thanks, you've illustrated my point perfectly."

Glad to be of service, but I'd rather use simple voice commands to control a portable device. My cellphone has the ability to dial by voice, recognizing both names and numbers. It's not perfect, but it is usually faster than typing or searching for contacts.

Voice control and other methods are only infants compared to keyboards, but just like the keyboard improved from a mechanical device on a typewriter into a simple multi-function electronic device, other input technologies will improve.

I'm just looking forward to the day when the computer interfaces with my brain and provides all inputs so that I can just lie in some tube and experience the reality that the computer determines is best for me.

Re:Yeah good luck with that

[smallfries]

There is a more general lesson here. When someone comes out of the workwork and says: "Look I've invented a new type of input device" then generally it will be interesting regardless of how well it works. When someone tries to flog the same dead horse that hasn't worked for twenty years then you know that it will suffer all the same failures. I'm sure everyone in this crowd has used voice activated interfaces and knows just how much they suck.

When a business analyst / investment "consultant" starts hyping up marginal advances as revolutionary and talking about coming "paradigm shifts" then you know that the bullshit is in full flow. Accelerometer interfaces are nice, they do feel more natural - I worked on one for an educational games project seven years ago. But the key point that you've captured is they are intrusive. Until the accuracy is high enough that we can make a twitch interface they are not a replacement for the traditional tools of mouse and keyboard.

What really pissed me off about the article was the insistence that these interfaces were a "direct manipulation" of images on screen. No, if you reach in and move an image (somehow) then that would be direct manipulation. If your physical gestures are translated into screen motion by accelerometers rather than a mouse then it is still an indirect interface. It is at most a minor increment on the user interface technology that we have already, the term "Paradigm Shift" is thrown about with abandon by too many suits without a understanding of what it implies.

Re:Yeah good luck with that

[jay-za]

It will be on a phone.

Most of the PDA phones available these days (or at least the ones available in South Africa) come with a built in GPS unit, or at least a GPS extension module. Touchy-feely isn't that great an idea when you're driving, so voice becomes an important issue here.

I've never seen one crash. Then again, I haven't had much exposure to Windows Mobile.

As the proud owner of an iMate JAM, an iMate K-JAM and a Mio P550, as well as having a number of friends with older HTCs and other iMates, I can assure you that yes, they do crash. And more annoying than when they crash is when bluetooth, or WiFi suddenly stops working, or when the addressbook suddenly appears empty until you reboot. My presonal favourite is when the handwriting recognition goes for a loop (that's a memory problem, I eventually found out).

But what REALLY gets to me is when the device just quetly hangs once it's gone into standby mode. With the phones, it meant I silently dissapeared off the cell network until I realised there was a problem, which is usuazlly when I try to use the device.

I have some friends with the newer HTC phones, and they report that WM6 seems to be more stable, but a few of them have reported serious problems with battery life.

I think there are a lot of things we need to sort out with mobile devices before we look at redoing the interface.

Re:Yeah good luck with that

[vertinox]

The keyboard works, 100% of the time. Its easily understood. Its robust. It fails gracefully - you immediately see if you've made a mistake before submitting a command.

True, but they should be focusing on other methods of input.

This could be anything from the one handed keyboard, ear canal senor that detects tongue movement, or mouse cursor that follows eye movement.

Personally, I'd wouldn't mind having an electrode in my arm or back if it means I could use small muscle movements to input text and mouse movement but that might be a hard sell to the average joe.

Re:Yeah good luck with that

[dkf]

You don't want to start arguing with your PocketPC about traffic and directions: No, I said Springfield, not Slingblade! *crash* It could be worse, much much worse: No, I said goats, not goatse! My eyes!

Re:Yeah good luck with that

[MacDork]

you immediately see if you've made a mistake before submitting a command.

Well, the problem with that is that you have to look at the darned device to do anything. Speaker independent voice recognition works quite well already on a Nokia N95. You hold a button, speak a name from your address book, and it not only displays and speaks what it thinks you want for confirmation, but it also has a list of next best guesses. You're not going Captain Picard with the thing, but it works well with minimal input. In noisy areas, just hold it close and speak up. You can't say that with most "smart" phones like iPhone and it doesn't demand your eyeballs if, for instance, you really need to place a call while driving. I use it all the time in preference to the keyboard because it beats flipping through the hundreds of address book entries in my phone. I like that direction in UI and hope we continue to see more of it rather than dwell on how glossy and cool our phones look.

Re:Yeah good luck with that

[Fred_A]

"Thanks, you've illustrated my point perfectly."

Glad to be of service, but I'd rather use simple voice commands to control a portable device. My cellphone has the ability to dial by voice, recognizing both names and numbers. It's not perfect, but it is usually faster than typing or searching for contacts.

Voice control and other methods are only infants compared to keyboards, but just like the keyboard improved from a mechanical device on a typewriter into a simple multi-function electronic device, other input technologies will improve.

I'm just looking forward to the day when the computer interfaces with my brain and provides all inputs so that I can just lie in some tube and experience the reality that the computer determines is best for me.

Where did I read
"Text-based interfaces have proven that most users can't read.
Graphic interfaces have proven that most users can't understand abstractions.
Mind reading interfaces will prove that most users can't think."

I have little doubt that it will happen that way.

Re:Yeah good luck with that

[OMNIpotusCOM]

I type over 70 WPM using what you term muscle twitches. With a very small amount of training I can use voice recognition software at over 160 WPM and it doesn't involve strain (other than the mental "strain" of enunciating). Aside from that, nothing is ever misspelled (homonyms and other nuances are all you must worry about).

People talk all day (ask my mother-in-law) without losing their voice or straining any muscles, but have you ever typed literally all day? It is unreasonable to expect someone to type as fast as they can dictate with the same amount of training in each.

On Vista saying "open notepad" is much faster than trying to remember where it is buried on the menu. People can pick up a mic with a list of key words in front of them and more easily use the computer than they could with a mouse. Other just touching what they want instead of determining the difference between left-click, right-click, double-click, drag, etc... This is the reason that programming languages that read closer to English are usually more popular, they're simply easier to pick up and understand. Nobody wants to remember syntax.

Maybe you shouldn't talk about things you have no experience in, let alone try to make analogies that bare no relevance to the discussion. Maybe your closed-mindedness is the reason that interfaces haven't changed much, but I'm willing to bet that you will get on your Iphone and call all your friends to discuss how stupid this poster named OMNI-something was on /. ... or if not you'll sure send them a text message using T9 instead of just pounding out each letter individually. Tell Fisher Price about that.

Re:Yeah good luck with that

[arth1]

On Vista saying "open notepad" is much faster than trying to remember where it is buried on the menu.

In Windows 2000, it isn't. What's changed is added bells and whistles to the start menu, as well as an artificial delay (presumably to help those who aren't good at quickly and correctly moving the pointer). With Windows XP, the "dynamic" menu was also introduced, making the menu in its default setting hide what you haven't used recently, and at the same time preventing any kind of spatial memorization of where to find things -- it can and will change. With Windows Vista, there's a huge big mess of trying to replace menus with predictive breadcrumbs (yet another way to prevent spatial memory), and some of these design choices have even hit the innocent start menu. To the point that it now /is/ very slow.
That doesn't mean having menus is the slow choice.

And it's a hell of a lot faster than repeating yourself multiple times, or having to use a menu /anyhow/, because it's too noisy for voice control where you are, or you have to be quiet.

What's needed, IMO, is a simplification of the UI, focusing on simplicity and consistency, and not done by trying to second-guess the user or provide a more "natural feeling". Saying "Enhance 224 to 176. Stop" might work in a movie, but in real life, it's by far easier to drag a mouse box over an area.

Anecdotes have it that the tree most common words said on voice operated telephone menus are "no", "dammit" and "operator".

Regards,
--
*Art

Re:Yeah good luck with that

[jay-za]

That sounds like a feature, unless you're one of those people who grows desperately frantic at the notion that you might be unreachable at any single point in your life.

You missed option (c). One of those people who's BOSS grows desperately frantic at the notion that I might be unreachable at any single point in my life. In those days (a full 2 years back :-) I was a technical contractor, and if my boss didn't know where I was for 30 minutes he'd phone and ask.

It would be even worse these days, though. As IS manager I'm responsible for everything that goes wrong at the office (a part of the job description I missed during the interview stage), and I would rather have my techs contact me when there's a problem, than walk into an ambush the next morning when I didn't know something had happenned. It may sound inane, but a simple problem like "Director X' home ADSL stopped working" can get escalated to "all Internet traffic, including traffic to that new and important project, dropped for half the day yesterday and no-one knew what to do about it" if I'm no able to babysit the problem, smooth the egos and make sure that no one over reacts.

Anyone know of a senior position in the IS industry where that isn't the case and I'll be glad to submit my CV. It's actually situations like that one that have made we consider giving up computers and taking up farming. The hours are easier and it's more predictable.

(That last comment was humour. It's funny, laugh. But don't mod me +1 funny, choose something else, I need the karma ;-)

Re:Yeah good luck with that

[Metasquares]

I realize you were joking, but as the poster said, you illustrated his point: let's say you wanted to dictate that response. Those aren't words, so how do you do it?

"T". "y". "e". "s"....

*5 minutes later*

"o". "f". "o". "Period".

No matter how fast the system responds, you can probably type the letters faster than you can dictate them. Similar things would happen when dealing with non-natural languages, such as programming languages. Can you imagine trying to dictate a regular expression? :)

A voice is a wonderful thing, but we should probably acknowledge that it's not always the most appropriate input method for the job. In some scenarios, such as writing a lengthy Word document or transcribing meeting minutes, dictation offers great promise (if we can ensure a high degree of accuracy), but it is virtually useless in others.

Re:Yeah good luck with that

[Dun Malg]

As the proud owner of an iMate JAM, an iMate K-JAM and a Mio P550, as well as having a number of friends with older HTCs and other iMates, I can assure you that yes, they do crash... I have some friends with the newer HTC phones, and they report that WM6 seems to be more stable, but a few of them have reported serious problems with battery life. Indeed, I used to use a Mio A701 (WM5) and it was a bloody nightmare. It required a reboot every 6-8 hours because the radio driver would quietly crap out (no error message, the phone would simply no longer receive or place calls!). I've had much better luck with my HTC TYTN II/Kaiser (an AT&T Tilt I reflashed with the HTC Kaiser ROM) with WM6. Battery life was indeed an issue at first, but if you're a dork like me you can try different radio ROMs until you find the one that lasts longest with your hardware. I recently flashed it to WM6.1 (requiring a 6.1 compatible radio ROM) and battery life is even better. As much as I'd love a non-Microsoft OS on my phone, I have to admit that WM6 is perfectly serviceable for my needs. It's relatively open so there's lots of software and hacks for it, and the phones with the features I require* all run it.

* slide out keyboard, GPS, touch screen, simple USB laptop tethering, and HSDPA, so don't point that stupid toy "1 out of 5 ain't bad" iPhone at me!

Visually impaired ignored?

[cojsl]

Hopefully "they" also develop good image to speech technologies, or are they forgetting that there are many visually impaired Internet users?

Doesn't bother me

[nizo]

As long as the extra flashy junk doesn't impede my ability to get useful information from a website, I will be fine with it. There have been so many sites that don't seem to understand this though (yahoo maps is a great example, among many many sites. The original "low bandwidth" version is still more useful than their "new bling improved" version, even over a high speed connection). Ebay is headed down the path of "bling overload" too. What bothers me is when a site adds rotating blinking things without considering, "what improvements does this give us or the user trying to use our website?"

Not only that

[dreamchaser]

I certainly don't want to be on a bus or plane with dozens of people all yakking commands to their devices, nor do I necessarily want to display to the world what commands I am giving to my device. Voice control is nice in certain circumstances, but until they give me a direct neural interface I want keys and/or stylus and/or cursor control and input options.

Voice is too slow

[Mprx]

I can think faster than I can speak unambiguous commands. Using a combination of keyboard shortcuts, extended mouse buttons and mouse gestures I can browse fast enough that the bottleneck is almost always reading comprehension. This is also much less tiring than speaking. A better solution might be a combination of eye tracking and brainwave monitoring, but that's still far too unreliable.

Interaction Language...

[krahd]

"'I've wondered for a long time why the computer interface hasn't changed from 20 years ago,'
OK, playing a little devil's advocate here. Perhaps the building bricks of computer interfaces and their basic interaction mechanics haven't changed because they are all right as they are now.
We have developed an interaction language that allows us to express interaction proposals and allows the users to understand those proposals and, therefore, to interact successfully with our systems. Why should we change that if it is working?

Change for change's sake, when we have an established language does not sound sound... I don't see no one complaining that we've been calling chairs "chairs" for so many years...

Yeah, okay, sure... You go first.

[JonTurner]

Talk is cheap. All this balderdash about next-gen interfaces, 3D, voice control, blah-blah-blah and how your great ideas will revolutionize the industry. Well, let's see it! How about some examples? The windowed GUI was an obvious quantum improvement for the vast majority of computer users (yes, I realize that on /. command line is king) but there has been no movement forward for nearly 20 years. Most importantly, the GUI window paradigm worked well. Let's see your prototypes rather than just more "big ideas" or is this simply a rehash of the "one day we'll have flying cars" speech, applied to computers?

I have to admit that I didn't agree with his ideas, but Jef Raskin, RIP, (original concept for Macintosh, "Swyft", "Canon Cat") was one of the few designers who was brave enough to take a clean-slate approach to interface design and then *implement* it to see if the ideas stood up to real-world use.

Re:Yeah, okay, sure... You go first.

[Haeleth]

The windowed GUI was an obvious quantum improvement for the vast majority of computer users (yes, I realize that on /. command line is king)

Even command-line users pretty much all run their terminals under a windowing system these days. Even if they use traditional editors like emacs and vi, most people default to using versions of those that take advantage of the features that GUI environments provide. And how many people do you think browse Slashdot from the command-line? Methinks the number is small.

So I think it's safe to say that the number of people who do not see any benefit from graphical windowed environments is infinitesimally small, even among hardcore *nix hacker types.

but there has been no movement forward for nearly 20 years.

How old are mouse gestures, out of interest? Most people who use them seem to think they're a step forward, and they've only been a mainstream concept for a few years, though I'm sure they've existed for far longer than that as a research concept or whatever.

voice control

[theheadlessrabbit]

voice recognition as it is today is painful.

"Computer, start, programs, Mozilla, fire fox , double you, double you, double you, dot, google, dot, com, search field, violent, asian, porn. I'm feeling lucky. click"

its a slow, painful, annoying as hell process that brings you back to the keyboard and mouse once the novelty has worn off, and only leaves the user feeling ripped off for wasting so much money on a fancy new inferior interface.

voice recognition won't be useful until it is intelligent. I should only have to say "Computer, google porn" and get my results. I shouldn't have to explain to my computer step by step how to open a freaken browser.

Re:voice control

[calebt3]

And then there is entering /.'s URL.

Here's an exercise

[Anonymous Coward]

Here's an exercise for those who believe voice commands are the way to go for small electronics. Every time you use your cell phone, iPod, PDA or GPS, say each command out loud before entering it. See if you can keep this up for a full day.

I prefer cross-platform standards.

[Haeleth]

Piclens looks cool and all, but it's just a proprietary program (like Google Earth, really) that happens to run in a web browser.

Want to use it on Linux? Sorry, you're out of luck, it's Win/Mac only for now; they say there'll be a Linux port one day; but as this is a proprietary technology, you won't get Linux support until they deign to implement it.

Want to use it with Opera? Sorry, you're out of luck, it's IE/Safari/Firefox only for now; and it will probably remain so, as they say they're not interested in supporting minority browsers; and as it's a proprietary technology, Opera can't add their own support for it.

Want to use it on an iPhone? Sorry...

This is not a step forward.

We'll see

[owlman17]

I don't know. In the 80s, back in the days of MS-DOS, I vowed never to switch from a CLI. A GUI (on a regular PC) was not only slow as molasses, I could think and type faster on a keyboard than use those new-fangled things called mice. I bought one just for the heck of it. It came with a primitive paint program and a TSR for shortcuts. I figured it'd have a niche but it would never hit mainstream. I wasn't the only one who felt that way. There's a lot of skepticism judging from the posts so far, but who knows? Resistance is normal I guess at the start. We'll have to wait and see.

Re:We'll see

[Steve001]

I think one of the hinderances to practical voice recognition has been the telephone paradigm (described in the book "Being Digital" by Nicholas Negroponte) where the computer is supposed to understand anything that anyone says at any time. What might work for voice recognition is for the user to have a custom chip that will allow a device to be configured to understand that specific user. Move the chip to a new device and that device will understand you perfectly.

What might also work is if the user trains himself/herself to speak in a way that the computer can consistently recognize, much like the user of Palm's Graffiti handwriting system learned to write in a way that the PDA could consistenly understand. With training, speaking that would could become second nature, much like typing has become for many users.

He wondered for a long time why...

[TransEurope]

It's because it works like it is. And the "new" ways of controlling aren't advantages, they are just ways of fixing the disadvantages of small displays and small devices lacking (working!) methods of cotrolling like mouses, trackballs and so on.

more like a tool

[radarsat1]

'People should think of a computer interface less as a tool and more as a extension of themselves or as extension of their mind.'

I wish people could learn to think of their computers more as "just a tool". Half the time I see people having problems with computer usage, it's because they're expecting the thing to read their mind. I have to explain to them just how dumb a computer is, and that you really have to tell it what to do because it's just a machine.

(The other half, of course, is due to shitty software.)

Baby steps first, then worry about how to best run

[dissy]

Many technologists now believe that hunting and pecking on the tiny keyboards of cellphones and P.D.A.'s will quickly give way to voice commands that will return map, text and other data displayed visually on small screens." Despite the fact most of us are extremely faster at typing than 'hunt and pecking', even the slowest hunt and pecker is going to be exponentially more accurate at input with a keyboard than even the best voice recognition software in existence today.

Voice recognition still sucks badly, even after a lot of time investment into it.
Maybe if someone got around to fixing that somehow, then we would consider, you know, using it.
I'm not at all suggesting we give up that line of research, just suggesting we put the horse before the cart here.

Or at least don't lie and say "will quickly give way to voice commands" and call it what it is. Those people want it to happen, and there is nothing wrong with that! Each tech has people that would prefer it over others. To each their own!
But to out right lie and say that it will happen 'quickly' is just embarrassing for your career as a technologist.

That explains it!

[Phoenix666]

The other day I overheard my neighbor two cubes over say the following in syncopated fashion: "teens," "threesome," "bukkake."

Simple answer

[glwtta]

I've wondered for a long time why the computer interface hasn't changed from 20 years ago

Because it works.

Whereas all the attempts at shifting the paradigms to an extension of your soul (or whatever), just result in unusable exercises in masturbation (and not the kind the internet was invented for).

Remember how Flash was going to be the future of the web? Yeah.

What about the icon-impaired?

[pongo000]

I'm icon-impaired. Seriously. My mind cannot make the subconscious connection between an icon or graphic and what said graphic is supposed to represent. Over the years, I've forced myself to recognize a floppy disk as "save," and a printer as "print". The rest mean nothing to me. When I use OpenOffice or any other graphic-intensive program, I must either (1) memorize various keyboard shortcuts, or (2) hover over the toolbar icons to find the one I want. For obvious reasons, my editor of choice is one that doesn't require me to decode icons. Nearly every graphical "decode" operation requires conscious thought as well as a process of elimination to narrow down the choices to a set of possibilities from which I will (hopefully) select the correct one. Many times I'm wrong.

Almost everything I do is on the CLI. I've been programming for nearly two decades, and I have no problems selecting textual tokens out of a field of similar-looking text. But give me a set of small, information-deprived graphics to decode, and I fall flat on my face.

I can't be alone in this. Surely others have this same cognitive disability.