My friend was trying to help his mom use the computer over the phone (he was at work or something). And he kept telling her to click the button, and sometimes it would work and sometimes it didn't. Finally, in exasperation he yelled, "Are you clicking the LEFT mouse button?!" She replied that she didn't even know there was a difference, and would just "click one of them" which would account for the roughly 50% success rate. :)

> [Double-clicking] was basically forced on the
> industry by Apple Computer when they elected to
> use a single button mouse on the original Mac in
> 1984.

* Microsoft develops two-button mouse in 1983
* Microsoft releases Windows 1.0 in 1985

Would you mind explaining how double-clicking was "forced on the industry" by Apple? I really don't think that the original Macintosh had such momentum that it prevented Microsoft from choosing what you seem to think is the superior interface.

Oh, right... well, maybe if Microsoft had actually done any original work on interfaces of their own, they might have been able to make those kinds of decisions.

The state of UI at this point is so confused w/r/t clicking that users don't know whether to click or double-click. If clicking doesn't *immediately* do something, they're apt to try double-clicking just to see if that was the appropriate action. As your anecdote points out, in applications that don't anticipate this, "unexpected result can occur."

I work with an app that uses single-click and double-click for separate actions. The double-click action is extremely expensive (it downloads a file, which can take 15+ seconds), but just as unfortunately, single-click is not always responsive. As you also note, the double-click interval is generous defined, so the action single-click/pause to see if anything has happened/single-click often ends up being a double-click, much to my annoyance.

It truly makes me want to sit the developer down so he (as I happen to know) could see what it's like to use the app.

"Oh, right... well, maybe if Microsoft had actually done any original work on interfaces of their own, they might have been able to make those kinds of decisions."

You're right; even though 2 button (at least!) mice are standard on Windows systems, we still need to double click to invoke activation. So that was a squandered opportunity for improvement.

More disturbing is the way that Apple *still* ships a single button mouse out of the box with their systems, and perpetuates it in their latest interfaces. The mistake is 15 years old; the time to start fixing it is, uh.. any day now. Click and hold if you agree with me!

I expect more from Apple than I do from Microsoft in this regard (eg, being ui innovators). The sad thing is, as Mike points out, the users are the ones that suffer.

" She replied that she didn't even know there was a difference, and would just 'click one of them' which would account for the roughly 50% success rate. :)"

That's really funny.

I do think there's a point of diminishing returns for mouse buttons, but that point is DEFINITELY more than one. Probably two, with the third wheel button as an optional bonus.

I doubt this particular user would have been any less perplexed with a single mouse button to 'choose' from. ;)

we have a huge application for gov of brazil, something like 3000 pages of aspx and we dont use this Enable/Disable button for operations in webpages now i?ve got alot troubles cos this, maybe you know how i can change my 3000 pages button control to do this, and listen the bad thing we dont use a specific custom control to button.. we use de default button control..

Yes i know i?m in bad days ..

Help us if you can..

As you point out, you'll have to replace those standard buttons with a custom ASP.NET server control. Won't be much work to write the control, but retrofitting the existing pages probably can't be automated..

roberto-

you could write a HTTP module that scans the HTML generated by every one of your pages, looks for <button> or <input type="submit"> tags, and modify the on click events accordingly; just make sure that you propogate any existing postback (__DoPostback()) logic that might already be wired up.

it's a bit expensive to do this at runtime, but in general, the CPU hours that you spend will almost deftinately be cheaper than the developer hours that you would spend to change every reference to button. ;)

erm, in that last post, my example tags got stripped out; instead of saying:

looks for or tags

it should say:

look for <button> or <input type="submit"> tags.

i guess thats what that "preview" button is for, eh?

You're saying a couple different things here:

1. Double-click is not intuitive
2. The user should have another button for that action

Correcting either of the two (from an OS level) would cause the user more trouble than leaving it alone. For whatever reason the double-click is here... and change hurts a user more than anything.

As far as Apple, they are UI innovators. The single-button mouse is more usable for the average user. Users like my mom who don't really understand the distinction between left and right (and would almost never use the right button anyway) are better served with a single method by which to express themselves to avoid confusion. And adding a second button to an apple mouse to indicate a double-click wouldn't be very nice to the www.apple.com/switch 'ing user either.

I think that's the most innovative part about every facet of Apple's UI. Less options equivalent power. That, and jewel buttons. Jewel buttons is perdy.

"1. Double-click is not intuitive
2. The user should have another button for that action"

Actually, I'm proposing that the entire UI be made single-click just like the Web.

Microsoft *did* introduce this in either Windows 98 or ME, I can't remember which. Open any explorer window then do Tools, Options, General, Click Items as Follows, Single-click to open an item.

Interesting idea [disabling buttons], though I've experienced weird behaviour when using this method in the past, presumably because when the button is disabled, it's value is not posted, and .NET needs this data to process a postback.

My solution was to use some javascript and the onsubmit handler, setting a flag when submit was pressed the first time, and then checking the flag on subsequent presses to verify whether the submit button had already been pressed, and if so, disallowing the subsequent submit.

Btw, just wanted to say - cool blog! Found it recently, after clicking through from an article on CodeProject (the one about custom exception handling). Wish you had full text in your RSS feed though :-)

Thanks!

I agree you probably wouldn't want to use the .Enabled property in ASP.NET to disable the button, but rather disable the button using pure Javascript. The fact that the button isn't enabled after the click shouldn't be relevant to any server side processing anyway...

You would think that disabling the button after it has been clicked (using button.disabled = true) shouldn't affect server-side processing, but having done some more testing, it does appear to interfere with postbacks.

Feel free to give me a shout by Email or on MSN, and I'll send you my test pages :-)

Hmmm, all interesting. The original blurb does seem to accuse non-IS folks of being stupid when choosing dble-click or not.

What about the idiot mouse driver that clicks twice, being too sensitive to any movement and has no interface to adjust it????? huh? You know detection of false clicks?

So now the fault is back in the lap of the designers!!! LOL. Idiots.

I have had the same problem with our new accounting package. It duplicates data. Having been a programmer I know the difference between the single and double clicks. So I am positive that I am not double clicking. The only thing I can think of is that my Radio mouse is sending double clicks when I single click. I changed the mouse to normal mouse today and am eager to find out if this is the problem. Have you had any cases like this?

Thanks
Sai

Have you considered the possibility that the idea of double-click <i>is</i> a good idea?

This is not information, this is a rant, man.

I don't buy that having a two-button mouse in 1984 would have replaced double-click. In 10+ years of multi-button mice it hasn't; and even Windows 98 went for a different model than left-click-select, right-click-open.

Perhaps that's because double-click claimed the Open function (or vice-versa), and right-click has mainly become a context menu. So right click is no longer free for Open.

I'm on the side of not having objects selected unexpectedly (e.g. by hovering). Leave single-left-click the way it is. Move Open to a context menu if you want ("you" meaning all developers and designers)

Douglas Engelbart, the inventor of the mouse, says mice should have many buttons:

https://www.engadget.com/entry/1234000840068219/

Jef Raskin, who created the original one-button Apple Macintosh mouse:

"What I did not see at the time is that multiple buttons on a mouse can work well if the buttons are labeled. If the Macintosh mouse had had multiple buttons, if the buttons had been permanently labeled, and if they had only been used for their intended function, a multiple mouse button might have been a better choice. A better mouse might have two buttons, marked Select and Activate, on top and on the side, a button activated by a squeezing action of the thumb. This last button would be marked Grab. Some mice at present have a scroll wheel on top that is used primarily for scrolling. Better still would be a small trackball in that location. The mouse would control the position of the cursor; the trackball could be used, for example, to manipulate objects or to make selections from menus that float with the cursor."

That's on page 209 of his book The Humane Interface (2000). The trackball comment is awfully prophetic considering the Apple Mighty Mouse introduced in 2005 has one exactly like that..

These issues always leave me feeling somewhat conflicted.

I have trouble enough accepting the fact that a certain proportion of the population do not have any learning skills (or think they are excused from the burden of learning that others accept as part of life) - so how much should we cater for them?

Not that things should be complex (I hate bad interface design as much as anyone and I'm always looking for ways to improve my skills in this area), it's just a lot of time gets spent (and sometimes good applications get ruined*) trying to cater for someone who, IMHO, insists that the world cater to their intellectual laziness.

On the other hand, no-one can force them to fit in, so if it turns out they are the majority, then we have to cater for them to a degree for everyone elses sake. This certainly hurts the head.

* I am thinking of examples where generally accepted interface principles that 90% of the world knows by heart have been sacrificed by decree from management to cater for the one dumb-dumb who refuses to learn them - the "I want the OK button to be called 'Fred' so I know that I should click it and to respond after being clicked on 5 times with the shift key down so I can't accidentally click it" type.

A good example - I had a user present me with a sketch of improvements to the interface to a system I wrote. She thought having Save, Cancel, New, and Find buttons was too complicated. She wanted a 'Go' button that did whichever action was appropriate at the time (someday I must get around to writing that generic class ReadUserMind.cs - I'll make a fortune). Yes, there may be a better way than what I did, but it was pretty standard, and I'm sure it wasn't worth redoing everything to cater for one person.

Of course it's hard to know where to go when you are told by their manager that you can't expect them to know how to use, or learn to use, a scroll bar or a menu.

"Not that things should be complex (I hate bad interface design as much as anyone and I'm always looking for ways to improve my skills in this area), it's just a lot of time gets spent (and sometimes good applications get ruined*) trying to cater for someone who, IMHO, insists that the world cater to their intellectual laziness."

This is not trying to cater for intellectual laziness. This is trying to cater for those that occasionaly use a computer. People who don't use it enough to warrent sitting down and learning. I shouldn't need to sit my nan or brother down and teach her that to open/activate something you need to do this tedious "double-clicking" that is awkward and hard when your hands dont have the flexibility they used too and your not used to a mouse. I shouldnt need to then explain that things are different on the web. Oh, and in this application too.

The whole idea of what should double-clicking is largely undefined or unknown. This is evident in system tray icons. Look now, which require a single click, which require a double click? Two of mine require a single click, one requires a double click, and one uses single click to bring up a menu, double to bring up a window. Confusing, no? I would say to extend and clarify the idea, double-clicking on an iconic representation to execute an action is usualy the most often use of double-clicking, so system-tray icons should be double-clicked. But then QuickLaunch apps are single click. But the system calander is double click. But start menu items are single click. But Explorer icons are douuble click. See how confusing it gets? There is an complicated structure behind what is single and double. Yes, I can tell you (aside from the systray icons) why each of those are what they are, but it's not immediatly obvious.

I have been playing around recently with alternate methods for an Explorer style file system as an isolated componant. The best I have come up with is Right-Clicking launches a file. Left-Click selects a file. Holding down left click brings up the context menu, or you can bring it up quicker by then pressing Right-Mouse. Or a dedicated middle button for bringing up the menu, though you can't rely on a 3-Button mouse. But the "hold" is as obscure as double clicking, only easier to execute.

A myrad of problems comes in trying to do away with double-clicking. Firstly, you have 3 functions. Firstly, select. Select must be consistant between text and files and other item representations (lists etc.). Secondly, execute/action. This must be consistant between UI elements such as buttons, scrollbars etc. and files. You now have two common actions, that of selecting items in listboxes, text etc. and UI elements to scroll etc. that are close together that now require you to switch between two buttons, which can be confusing and takes homing time. Unless you combine them. But you then have the file problem of selecting a file and executing it. The last is menu. Without combining two actions or other UI elements you have no hope of supporting 2-button mice.
I personaly am in favour of removing menu from the mouse altogether. Left click activates. Right click selects. If you have a three button mouse, the third button is the menu. Otherwise a keyboard button. Most have a menu key already, so there is a high chance they will have a third button or the key. But some other combination could also be used.

Interesting thread! Lots of good points raised.. a few things:

The submit button at the top of the page breaks a cardinal rule of UI design.. never change the label on a button when it is clicked! It just confuses the beejeebers out of the user; like 'where did my button go? It was there a second ago! What did it say again?? How do I get it back?? I want my button back!!! (runs from room screaming AAAIIIIEEEE!)

As for double click vs. single click, there are pros and cons to both.
Double-click
PROS
- safe in terms of accidentally launching an app
- well known behaviour and more consistently used in Win OS's which is around 90% of the worldwide market
-
CONS
- RSI and carpal tunnel, although incessant single clicking can have the same effect
- 2 step process; select and launch

Single Click
PROS
- efficient

CONS
- can easily launch apps by mistake
- click and go model used throughout a UI can be unforgiving to the user, and disorienting.
- not well understood in a software UI; a web interface is different - you are just flipping pages - software apps are usually for performing work tasks where a mistaken click can have consequences.

This is not an exhaustive list by any means, just off the top of my head..

one real life use case to consider; one OS introduced a single click UI in one version (not sure which one ; Mac or 98/Me/2000?) and users were confounded by it. They eventually changed it so that double-click was the default, although users can switch to single-click if they prefer. Basically a rollover is the the selection and the click is the launcher.

Windows XP (as 95/98) includes a one click mode: if you select Folder Settings from Windows Explorer / Tools menu, you have "Single click to open an item" in web browser style.

I tried to use it once, but I kept opening icons I just wanted to move along...

I know the double click is difficult for someone (I had an 80yrs old customer dragging everyday its icons close to trashcan trying to dblclick...), but I think it would be more difficult to have them learn how to use more than 2 button (I had another customer with these big fingers always clicking both buttons at once...)

I remember when I first encountered a computer with a GUI -- the Atari ST, with its GEM desktop. This must have been 1988. Being used to command interfaces like BASIC and DOS, this was totally unintuitive to me.

Selecting with the mouse, dragging and dropping, etc. were all intuitive. But I could not figure out how to launch programs using the mouse! Lo and behold, there was a section in the "Desktop Reference" manual that explained the "double click". I always thought it was asinine how deleting the entire contents of the disk (dragging the disk to the trash) was more intuitive than looking at its contents (by double-clicking its icon).

The ST's mouse had two buttons, but the right mouse button only opened the top menu bar, and you would release it to select a menu option.

Mouse chording or gestures would be great alternatives to the evil double-clik.
Mouse chording is pressing one button while holding the other. Opera, the browser I'm using right now, uses it to move back and forward in history. Using left click to select and left hold and right click to activate would be easy to learn and to use.

Mouse gestures, despite the author saying "What next? [...] Mouse Gestures?", are really handy, I open pages, go back, close, minimize and even jump one directory up in Opera with mouse gestures and they always work fine. Once you master them, it's even simpler than double clicking or opening context menus.

"Once you master them, it's even simpler than double clicking or opening context menus"

Try explaining to my dad how to perform a mouse gesture. ;)

3-button mice should have been the norm from the start - one button for "select", one for "open", and the third for "context menu". This would be both easy to use, and easy to discover.

I was learning some Microsoft application a few years ago. Reading the manual, I found that clicking on a button would elicit a common behavior, and double-clicking would trigger a less used action. However, that was not all. Triple-clicking the small button would trigger another variation of the tool! Needless to say, I could rarely hit the triple-click on the first try.

Triple-click MUST die!

"The whole idea of what should double-clicking is largely undefined or unknown. This is evident in system tray icons. Look now, which require a single click, which require a double click? ..."

What this is really about is that Windows has an inconsistent UI.

"More disturbing is the way that Apple *still* ships a single button mouse out of the box with their systems, and perpetuates it in their latest interfaces. The mistake is 15 years old; the time to start fixing it is, uh.. any day now. Click and hold if you agree with me!"

No, I'll prod the trackpad with two fingers... or maybe I'll press Cmd-Click, or was it Option-Click? No... Alt-click then?

My parents habitually double-click URLs, and despite telling them there's no need they continue. Why? Because it works. Surely the fact we can send one command (double-click) to our computers and have it be interpreted as another command (single click) is even more broken than mice with one button; it'd be like Cmd-Q usually meaning 'quit', but occasionally and with no obvious logic doing something else like close the current window.

Let's have a five button mouse, one for each finger:

Select, Activate, Grab, Open-In-New-Tab, Switch-Apps

Few years ago I started using Linux with KDE full time. At first, its 'single click to open' annoyed the hell out of me, especially because it felt like like it was there to dumb down the interface and make it more usable for beginners.

But anyway, I decided to stick with it for a few weeks, just to see if there was any way I could get used to it (after all, I got used (read: addicted) to vi key bindings and modes) . I can't really tell when that happened, but after maybe a month I booted Windows, and was genuinely confused (for about two seconds :>) when I clicked an icon on desktop and it just got selected, not opened. I thought to myself: '...damned Windows, even opening Windows Explorer hangs them...' :).
Now I configure all my boxes for single-click-open, both Windows and Linux. It's just more efficient and has no apparent drawbacks.

One more thing, my main way of browsing the file system is with Krusader, using the keyboard. Few times I used Konqueror, and some of my friends that were around, saw me single clicking through the directories and were just amazed how fast (it seemed to them) the process was. Some of them even asked me to configure their boxes for the same behaviour.

"More disturbing is the way that Apple *still* ships a single button mouse out of the box with their systems, and perpetuates it in their latest interfaces. The mistake is 15 years old; the time to start fixing it is, uh.. any day now. Click and hold if you agree with me!"

What?

Okay, I'll grant you that Apple still supports clicking-and-holding for contextual actions.

But every desktop Mac I've seen for the past several years has shipped with a Mighty Mouse (which isn't exactly the most popular mouse, but does have more than one "button"), and every trackpad-based laptop they ship (that, to be fair, does have a single physical button) comes enabled with the ability to right-click by using two fingers -- yes, this is just as "hidden" a feature as double-clicking, but the ability is there, and I've not yet encountered a Mac application that doesn't respond to right-clicking.

I agree wholeheartedly with your main point about double-clicking -- we're building a web application at work and we decided (against my wishes) to include javascript double-click event handlers as the main interaction with our system. Even our own testers sometimes get confused, and that should say something.

But please, there are plenty of valid reasons to trash the Mac and its UI. Let's not claim 10-year old facts as relevant.

Roberto: Sure, you can probably fix your buttons with some javascript. You'll need to add an onclick handler that finds all input tags with type=button or type=submit, and attaches a custom onclick handler to them.

You don't have to change all pages just to make this change, although it's the preferred approach because onclick handlers create delays in rendering.

I wouldn't disable the button with Javascript because it puts the onus of correctness on the client when the error is happening on the server. Usually it's a better idea to arrange things so the POST is idempotent.

Darius: I completely agree; keep the corrections completely transparent to the user. Have it so that, even if the user clicks the button twice, the script running on the server will ignore the second request. The user won't be able to tell that their client has just sent off two requests.

Ah, nice hint. This one goes right into the submit button template :)

"More disturbing is the way that Apple *still* ships a single button mouse out of the box with their systems, and perpetuates it in their latest interfaces. The mistake is 15 years old; the time to start fixing it is, uh.. any day now. Click and hold if you agree with me!"

Never used a Mighty Mouse huh?

Scratch my last comment as I just noticed the post date on this article. Really should be a warning for reddit users.

Well said. I agree the double-click is more of a hindrance than a help. And as for witnessing the damage first-hand, my mother, who actually uses a computer for a significant part of her job as a high school librarian still double-clicks things she shouldn't.

It's a habit that needs to be broken, and this is one area that I think the web is helping, by giving us some other options to think about.

Concerning MS's "implementation" of single-click interface, it was horrible and should not be considered an example of a proper interface. Its main problem? It was based on timed delays, something I have always been adamantly against in user interface. Timed delays only serve to artificially limit the productivity of a user and do not apply equally to all users--another reason, in fact, that double-click sucks.

Cheers.

While this sounds reasonable at first, I'm not sure it holds up empirically. I had to do UI design (and tech support) for a while, and I discovered that users tend to have much more trouble with right-click than double-click.

And I respectfully disagree that this is a strength of the web's UI. This is one of the things I find most frustrating about the web! How do you select some text that the webpage's author decided to make a link? It's really hard!

I find single click useful for short lists of properties when using nautilus(since it's possible to expand the little info showing files size etc to include things like perms, time created/modified or whatever), single clicking is rather common for me(since I always right click and choose an appropriate program to open things with anyway, I rarely want to use the default for anything).

That and I'm all too often using scripts to do tasks the easy way.

I agree above poster, I really appreciate the mouse chords and gestures of Opera. As with the double click, these are useful, but unintuitive. Neither should probably be used without providing a more intuitive alternative (as is done in Opera).

Can anyone please tell me how to start a software business.
I am a B tech. in Computer Science.