I find that double click is a great gesture that could be successfully used in graphical interfaces but for some reason has negative connotation on the web. Should it be avoided at all costs?
I think it would be okay for certain uses, particularly if you're considering having an object draggable or highlightable on a single click, then on a double click it could be openned. However I would strongly avoid anchor tags being doubleclicked, or anything that only responds to a double click.
Double-clicking on the web should be avoided because it goes against the general practice of single-clicking links, and would likely be confusing.
...double-click must die since it causes novice users great difficulties and since it conflicts with the single-click interaction style of the Web
If your application does a good job of emulating an operating system double-clicking might be a consideration, but even then I don't think it would be worth the potential downsides.
I think it's more about maintaining conventions and user's general knowledge of those conventions than anything else. Browser/website interaction has been fundamentally different to OS user interface interaction since basically the dawn of the browser.
In native UI's the convention has been to have a solid, non-clickable window area with UI elements akin to physical buttons, check boxes and such. Hovering may highlight an item, but is used mostly on button hover states and other minor hints. Most hover reactions merely change the cursor to something to indicate a different action may take place, such as the text caret cursor. The double click is synonymous with opening something, a file, a window from a menu item and such.
Website interaction is more like a giant wad of text. Double clicking typically selects a paragraph or line of text and html content, akin to a text editor. Clicking usually means a change in the webpage, either through hyper links or form submit buttons. Hovering therefore has been used to show the responsiveness to clicking of different elements on screen. Hyper links, drop down menus, buttons, etc.
I think it would be unwise from a user experience point of view to try to override the double click behaviour of the browser, however useful you might find the idea.
(edit) That is until your web application becomes so akin to a desktop app that it warrants the need for double click to be used. Google Docs Spreadsheets is a good example of where it's absolutely intuitive for the most part that double click would perform cell based interactions, just like Microsoft Excel and iWork's Numbers do. I forgot to take these into account in my initial answer, but I believe that if the user interaction is such that the UI feels like a desktop application in the way the UI responds to the mouse, then absolutely.
Avoidance -vs- Inappropriate?
It's not that
double-click should be avoided. It's simply that it has a narrow band of appropriate usage. There is a distinct difference between the two.
An interaction that should be avoided is one where alternatives should be actively sought out to avoid using it, such as with flash banner ads and blinking text.
An interaction that has narrow usage, is one that is useful in very specific niche circumstances. Take for example the oft frowned upon
popup window. The use case is where a user explicitly wants to open a new window of specific contents. This is often done as a
popout such as in Gmail.
So where does that leave us with
double-click can be used to trigger a secondary action that doesn't conflict with the primary action caused by a single
I'd take a wild guess that everyone participating in this particular question uses
double-click on websites regularly, possibly without even realizing it.
The first example I'll give is native to the browser:
Text Selection: If you want to select a single word, you
double-click on it and it gets highlighted. As far as implementation goes, many web-based wysiwyg editors override the browser's default text editing capabilities, and might need to re-implement this particular feature to work correctly.
I do the second unconsciously:
Full Screen: Go to any video on YouTube and
double-click the middle of it. What happens? The video goes
full-screen. It's not just YouTube, any video player worth its salt online will use double-click to full-screen the video. You might even notice that the first click pauses the video, and the second click un-pauses the video. The interruption is accepted because the video is changing size.
An application where double-click would make sense would be in grouped item selection. If you wanted to select one item, you could click once. If you wanted to select a bunch of items of the same type, you could
I too think it to be unwise in anything but a custom, backend system where you can actually 'train' users if you will.
When trying to implement it "into the wild" (publicly accessible web sites) you will quickly find out how many novice computer users think all clicks should be double clicks. For the users who understand the difference between clicks and doubel clicks you will face issues simply because the idea of double clicking for functionality
In my opinion, double-click should be avoided everywhere, not just on the Web, because people who are just beginning to learn how to use a computer sometimes double-click too slowly, causing two single clicks to be triggered in the code.
If you really need to use double-click, try to make it obvious that double-click can be used, for example, by showing a popup saying what the user can do, the first time the user uses your application.
Just my two cents. Ad@m
Double Clicking is not very intuitive for a lot of users these days. Not to mention in the coming generations. It the world of touch devices, single taps are also more intuitive than double taps.
However, double-clicking or double-tapping is a fantastic feature for power users. Image a time where you have a user that needs to access a secondary action. You could take them into some horrific secondary menu or toolbar. A double click/tap could be a major improvement, efficiency-wise, in the context of an interface.
The question you need to ask is "What is the most intuitive step for the user to take?". If that is a double click/tap, then go for it!
TiddlyWiki is for example a javasript Web App with double-click editing and keyboard shortcuts. It is widely accepted. So im not sticking with the one-click mantra here. Of course on simple web pages one should keep with it, but on web apps this rule of thumb doesnt make much sense to me.
It depends pretty much on the type of web app, is it orientated to the mass (e.g. nytimes app) or too small group because of its content/functionality (tiddly wiki)
Firefox is my standard browser because with mouse gestures and Tab click action editing i save alot of time. one click mantra sounds like a contradiction to me. As Einstein said "as simple as possible, but not simpler"
I had an experience where i instinctively tried to double click a link. When I was reading the docs for (Backbone.js and was browsing examples. I clicked the example links in the menu to the left and the page scrolled automaticaly to the example. But when I wanted to visit the example directly i instinctively double clicked the link in the menu, expecting to be transported of site to the example page.
The main problem with double clicks in a web app is that you will have to create special code to handle this on touch enabled devices.
You can tie together touches with clicks and touch move with drag but you will have a hard time connecting a double tap with a double click.
Drag and drop and regular click is easier to implement - https://mobiforge.com/design-development/touch-friendly-drag-and-drop
If your application is browser based, it will still be accessed on tablets and smartphones. Also, don't forget touch displays for PCs are on the rise.
If you want your application to have a friendly UI across devices maybe it's best to stay away from double click, or you will have to create different functionality for the various devices.
Correct me if i'm wrong, but the first thing I thought of was that if you use double-clicks, you'll probably have to let the system wait for a tiny bit after any single-click in anticipation of a double-click (assuming the single click performs an action, not just a highlight), which will create a general perception of slowness while using the system.