User Centered

Unless... you finally come to learn keyboard shortcuts and leave the mouse altogether. ^_~

That's true. But you do need a mouse for most web browsing. Spatial navigation works pretty well, but the mouse still rules for browsing. I am a big keyboard nav nut as well and have some very cool anecdotes related to the benefits of learning them

in general, I like to do as much with the mouse before I have to move to keys, and as much with keys before I move to the mouse. I just hate swapping back and forth all the time between the two.

Firefox is still bigger than Opera in file size yet has a lot less features

I will try to use the keyboard shortcuts when I can remember to, right click on the mouse I use a lot

I do like Opera's UI and Tamil's slim skin for Opera 10 is very nice way to slim it down

Eddie! A brilliant proclamation.

I can't tell you how many times I've "back swiped" in Internet Explorer without thinking (or tried to close or go forward). It is one of the single most brilliant innovations Opera may be credited with... and it's a big one.

Excellent comparing the latest JS speed obsession fad, too.

Yum

I'd really like to see an eye-tracking/morea-esque/usability study done comparing an experience mouse gesture user against a non user to get a tangible metric of "locus of attention" as it pertains to the UI.

Seems like something Opera should do, Kenneth Maage?

Anonymous writes:

Seriously. Get a mouse with a dedicated button for "Back" and learn the basic keyboard commands like Ctrl-W and Ctrl-T. You'll never again have to think about the unintuitive "mouse gestures". Works universally, and equally brilliantly on every browser and other apps as well.

I've updated my post a little bit as a result of this previous comment.

I do use basic keyboard commands. In fact, I use advance keyboard commands. In fact, fact, my latest post in the forums is about the lack of keyboard access in Opera's dialog boxes. I love keyboard control. But admit it- most of the time browsing requires a mouse, but browser UI/chrome commands can be done with the keyboard. So switching back and forth between the two is a minor inconvenience, I admit, but that's the whole point of this article- milliseconds of rendering time is seen as what makes or breaks a browser, but interface interaction, the biggest bottleneck of use, is never factored in. Going back to the posting here- the "Locus of Attention" switching back and forth from mouse to keyboard is something that I'm sure can be measured in terms of milliseconds or whatever metrics are used typical browser "speed" comparisons. Browsers are fast these days. The biggest bottleneck is you physically manipulating the UI.

Dedicated back button? okay, but what about a close page button? Refresh? new tab? ....I don't use a lot of gestures, but those handful there are CRUCIAL to browsing and get used constantly- but I don't want dedicated buttons all over my mouse for those.

....and -I already don't think about mouse gestures. Sure from a usability perspective they are difficult to "discover," but what's more intuitive than muscle memory? We've been doing that since we were born. I think you're thinking more about "discoverability" than you are intuitiveness.

For what it's worth, I also do the same thing when I have the keyboard in my hand. For instance, if I'm filling out a form online, and even after I submit it, I will stick with the keyboard until it's more beneficial to use the mouse again (i'll use Opera's spatial navigation)

...perfect example of muscle memory and efficiency of UI in play- when I open a new tab using mouse gestures, I've noticed that I immediately (before the new tab is finished loading) I'm already moving my mouse over to the speed dial I'm thinking of.

Excellent follow-up E. I consistently find myself using mouse gestures because the vast majority of mice aren't "5 button mice" -- and why should I have to spend that money anyway when the software can do it and offer more.

I consistently use duplicate tab (including browsing history of that tab), open new tab, close tab, back, forward, and even hide opera. I'm not sure where the "duplicate tab" mouse button would go.

Yum

Right-

I left it out for the sake of making it sound like I'm overboard with gestures, but I also have a gesture to tile windows vertically that I use rarely but I like quite a bit...and a fit-to-width gesture that I usually use along with the tile.

Anonymous writes:

I fully agree with you that the Javascript speed is minor thing while usability and features are major things for choosing a browser. However, I can't quite understand your love for mouse gestures, especially making them the one main thing about browsers, to the point of completely discarding browsers without them.

People usually have two hands and at least right-handed people can easily use the left hand to do the basic tasks like "close tab", "new tab" as well as switching between the tabs, all at the same time as using the mouse in the right hand. As I said, this works universally in every existing browser and also most other apps as well.

I just think your approach to the problem is as flawed as is the whole "mouse gestures" feature for solving the problem of "locus of attention". Since there already exists (and has existed for a very long time) UI design standards for all the tasks you need, inventing new - proprietary - solutions is counter-intuitive and plain wrong approach, from the point of software design.

From the point of the end user, the right approach would be to learn to use the universal commands working on every software. Of course I can understand that some users might find the mouse gestures usable and intuitive, but being proprietary, that unnecessarily limits the number of possibilities - just like you are limiting your choice of browsers to only those with the given proprietary feature.

Using the universal commands doesn't artificially limit the number of possibilities, so that you can make the choice based on other, more important issues. Maybe even the Javascript speed.

Following your logic the mouse itself is a proprietary solution that is "counter-intuitive and plain wrong". Most of what can be done with a mouse can be done with universal commands on a keyboard that don't "artificially limit the number of possibilities" and are standardized.

The introduction of the "mouse" broke all the interface standards and even the most basic interaction with computers... no matter that it also had the nice side-effect of making computers tremendously easier to use.

You cannot simply ignore the efficiency of a single-handed operation because you view the "right" way as two-handed "universal commands working on every software". The "right" way has always been the most efficient way (how do I get from Point A to Point B with the least amount of energy?)

Perceived efficiency guides the evolution of the computer interface. [Further, I'd argue we are very good at naturally choosing the legitimately most efficient method of doing something -- eventually.]

It would seem you are making a number of assumptions:

1. There is a set of defined methods for human interfacing with computers which may not be deviated from, ever. (the fact that we aren't using punch cards still, demonstrates how this is incorrect)
   
2. That there should be only one way for interfacing with a web browser.
   
3. That software designers should not innovate and should constrain end-users to meet the demands of the interface as opposed to software designers (and hardware designers for that matter) meeting the needs of the end-user.
   

That said, I like your idea in terms of an "ideal world". I'd love for mouse-gestures to be internationally standardized and used the same way throughout all computer OSes... However, we know that's not how it works. "Proprietary" systems always precede standardized systems.

Mouse gestures are clearly not for "amateur" surfers, they are for those who spend a lot of time on the Web so that the habitual nature of muscle memory (think of any sport) then allows for faster pathways to form in the brain for efficient navigation.

In other words, the methodology you are proposing would be akin to having to visually locate the gas pedal in your car every time you went to use it. Instead, after the period of habituation, we rely on muscle memory so that we don't have to "think" about which foot we use to accelerate. Instead, we simply "think" accelerate and our mind & body does the rest.

It all comes back to muscle memory (it's the difference between hunt-and-peckers and 126wpm).

Finally, a personal case-study.

Mouse-gestures were the feature of Opera that sold me on the web browser. Hands-down! (ha) It may be a silly reason but I lived in a very cold climate (and heat was expensive) and I'd often be in a blanket while using a computer (wrapped around my whole body with just my hands and face sticking out). I'd often switch hands so I could keep would hand under my blanket while the other used the computer.

What possible gain is there in forcing me to use your 'standardized interfaces', either I use two-hands or I am forced to change my locus of attention? Mouse-gestures require neither.

I hope I didn't hit this one too-hard but the elegance and efficiency of mouse-gestures cannot be understated.

Yum

My proclamation is rooted in that browser are, in most cases, pretty much the same these days. They are all relatively secure, I'm relatively able to navigate the brave internet waters while avoiding phising, they all have tabs and bookmarks and bells and whistles. So the differentiator is (as you would expect to find in a user centered/usability blog) how I use the interface.

I'm all in favor of consistency and not having to "relearn" every time I get a new piece of software or website. It's what I do for a living. Opera has changed their keyboard mappings over the years to be more in line with other browsers, and I have appreciated and applauded that. But exactly like WillYum points out in his numbered list, I also find tremendous value in being able to do different things with a UI- tailor to particular audiences. I think it's perfectly acceptable to offer advance features (and yes, I'm calling mouse gestures an advance feature) that expand upon the standard ones. We should be consistent in our software development, but not constrained by it- otherwise there would be no innovation. Or hell, even deviate from standards (UI standards, not web standards) and accept the consequences of small marketshare if you can come up with an interface that matches a particular niche of users (some would argue this is the reality of Opera).

...but OPera does stay on par with keyboard shortcuts and expected behavior found in other browsers (granted the skin is a departure). Since Opera support all the standard, expected keyboard shortcuts, what difference does it make if the offer mouse gestures as well? I've argued several times in the past that they don't "get in the way" of normal browsing, in fact, I think they are rather hidden a bit too well for beginning users to discover.

The bottom line is that mouse gestures match the way I want to browse best. Let's go back an look at all the times in this post/comment thread alone that I've said I do use keyboard shortcuts OFTEN. I do. I like them. They are very nice. BUT- mouse gestures are a very useful differentiator for me- so much that I can't stand to use a browser without them.

...and in closing; going back to WillYum's comments on one handed operation:
I currently have a coffee cup in my hand I browse the morning news feeds- how easy is it that I don't have to even take my hand off the cup in order to navigate around the UI? I mean what better illustration is there than that?

For "back", I press backspace. Works not only in browsers but image viewers and probably some other programs. For "forwards" or "continue" many programs use the space bar (e.g. image viewers, news readers, and many media players). Unfortunately not browsers though.

I think the important thing is having a common user interface for all programs (what used to be pushed as the CUA) so that no matter what program you're using, the same operation can be done in the same way without having to learn some idiosyncracy of a specific program. Of course every program has unique things it does. But most programs also have many things in common that they do. What we need is for positive innovations in interactivity to catch on across software generally, if they remain specific to one program I feel they're not much use. At present mouse gestures are just not widely adopted and so I don't find them of much use.

In some programs in fact they're a real hassle. I use an image viewer, XnView, which has mouse gestures that can't be disabled. When I try to scroll an image to the left, 50% of the time it is interpreted as a mouse gesture and causes the next image to load. When I try to scroll up, often it dumps me back to the thumbnail view for the same reason. This has made me wary of gestures, and not wanting an action such as selecting text by dragging the mouse to be interpreted as a gesture, I decided to disable gestures in Opera.

infinity-1

What we need is for positive innovations in interactivity to catch on across software generally

100% agreed. Consistency is a double edge sword though, it's hard to innovate (UI interactions) if you're always doing the same thing over and over again right?

if they remain specific to one program I feel they're not much use. At present mouse gestures are just not widely adopted and so I don't find them of much use.

I don't know, why aren't they much use? Consistency is great (as I mentioned, I do usability for a living, I get that), but a company (Opera) should be able to find the right balance of consistency with other software (browsers), while still being allowed to take risks and offer alternatives. Going back to your first part of your quote, they need to be allowed the room to innovate.

Mouse Gestures may not be widely adopted (by general users) but they are very widely adopted by me. So much that I cannot use a browser (comfortably) without them. I realize that I'm in the minority by far... but hey, I made this proclamation so that you all (and Opera) will hear this not-very-popular feature elevated to the status of the primary UI differentiator for me. People are different- we need a common standard (consistent UI) with which to operate, but we also benefit from particular nuances that match the way would like to work.

I view it like regional dialects of a language (it's a poor metaphor, but all I could come up with right now) we all have a foundational understanding of the language, but some people feel more comfortable and can communicate more effectively using shortcuts and jargon that their community understands.

XnView- I've never used it, but forcing gestures on a user is not the way to go. This sounds like a poor implementation of the feature, but I think it's not in any way an indictment of gestures in general -but your wariness of them as a result is really a shame, I wish you would have had a better experience. Or even no experience, I'm not saying EVERYONE benefits from using mouse gestures, just that I really, really do.

...and what's wrong with a proprietary feature or a feature that only exists in one piece of software?

First off, it's not. I use mouse gestures in Firefox, I've seen similar Windows and Mac based tools that allow gestures, but quite frankly, they don't really fit that well outside of browser. That's fine. They don't have to be. I don't get much use out of them, others may, that's fine.

But what about "Speed Dial" as browser UI element? Is that not a good idea either because it's not implemented (widely) in other browsers? What about extensions in Firefox? Is that a bad idea because no other browsers have an extension capability?

Firefox is just as usable without extensions, and the feature doesn't get in the way. It adds tremendous value to those who wish to exercise it. I feel the same about mouse gestures.

I'm going to echo Eddie on this one. I want to be clear that I'm not advocating mouse gestures be the predominant UI interaction.

A racecar driver's interface with his car is different than that of a regular driver but the basics remain the same.

I want to drive faster on the web....

Yum

Originally posted by Eddie_Lopez:

I've noticed that I immediately (before the new tab is finished loading) I'm already moving my mouse over to the speed dial I'm thinking of.

I just tested something that you might find interesting...

Mouse Gestures lengths don't seem to have have any limits. You can string together six movements if you want. For your speed dial entry two, you could move down to start with the normal gesture for new tab, then "draw" the number 2
GestureDown,GestureRight,GestureDown,GestureRight Go to speed dial, 2

Drawing the number 8 would be a simple:
left, down, right, down, left, up, right, up

(ha ha, so maybe 8 isn't that easy...)