Mozilla’s Tab Candy is the first step to sweeter browsing
July 29, 2010
Tabbed browsing has arguably had a significant impact on the way that people use the Web, but the feature hasn’t really scaled to accommodate the increasing complexity of the average surfing session. The existing tab management and overflow handling mechanisms that are present in modern browsers are dated and suffer from some fundamental limitations that significantly detract from user productivity.
As more software shifts into the cloud and users increase their reliance on the browser for daily computing tasks, browser tabs will have to evolve from a primitive mechanism for switching between documents into a full-blown task management system. The mainstream browser vendors have been slow to address this issue and haven’t applied much innovation to the problem over the past few years. Mozilla has stepped up to plate and is aiming to hit the ball out of the park with some unique and truly compelling improvements to the tab concept.
Mozilla’s experimental Tab Candy project, which is led by talented designer Aza Raskin, offers a simple and intuitive new twist on tab management. It allows users to visually manage tabs by organizing them into spatial groups. It’s far from being a complete solution to tab overflow, but it’s a very good step in the right direction.
Mozilla has made available some experimental prerelease builds of Firefox 4 that have the Tab Candy enabled. We tested this preview version ourselves to get a hands-on look at the new feature. On the surface, the only major noticeable difference is an icon with black squares that appears in the tab bar. When you click the icon, the Tab Candy mode will be activated. The browser will show you a thumbnail view of all of your tabs in rectangles that represent groups. You can drag a tab from one group to another or drag it out into the field to create a new group.
When you click a thumbnail, the browser will activate that tab and close the Tab Candy view. During regular browsing, the tab bar in the window will only show the tabs from the group that is currently active. This makes it easy to treat tab groups like projects and easily switch from one tab context to another.
These features are just the start of what Mozilla has planned for Tab Candy. In a demo video that highlights some ideas for future features, Raskin discusses the possibility of enabling simple tab sharing through the Tab Candy interface and providing extensibility hooks that would enable third-party add-ons to augment Tab Candy with their own contextually relevant features.
Tab Candy is an impressive first step, but there are still a lot of unsolved tab management challenges that need to be addressed. The Tab Candy interface won’t fully resolve the problem of an overflowing tab bar, because there are still likely to be cases where individual tab groups have more items than the regular tab bar can cleanly accommodate. Having to scroll back and forth to find a tab is frustrating.
Tab Candy’s spatial view will help to simplify high-level tab management, but the downside is that it fragments the user experience by disconnecting the tab management interface from the regular browsing interface. It would be good to have a separate way for users to optionally view the complete stack of tabs from all groups alongside the actual content of the active page.
Mozilla has reduced the challenge of finding a specific tab by introducing a switch-to-tab feature in the AwesomeBar, but that doesn’t help unless you remember the title of the page that you are looking for. The popular Tree Style Tabs add-on offers an elegant way to further simplify tab management—one that could potentially work well with the Tab Candy concepts and shore up some of the weak points.
The Tree Style Tab add-on allows users to see all of their tabs in a nested hierarchy in a sidebar. It presents tabs as a tree of collapsible nodes, which makes it easy to hide and show sets of nested tabs based on which ones are relevant to your current activity.
I think that something like Tree Style Tabs should be added as a sidebar, giving the user the ability to toggle between the regular horizontal tab bar and the richer tree view when the tab count becomes overwhelming. It could also potentially be adapted with a filtering mechanism so that the user can decide if it should show tabs from all of their Tab Candy groups or just the active group. The groups could be presented as tree nodes.
I think that a vertical interface is really the key to bringing saner overflow handling to the tab bar. Raskin is no stranger to this notion, and experimented with the idea of a vertical sidebar in some mockups last year.
In the demo video, Raskin suggests that Tab Candy users might want to rely on groups to manage the tabs that they intend to read later. This approach makes sense, but it might not be sustainable in the long term. I know that I’d end up with a ton of groups that I haven’t looked at in a while cluttering up my Tab Candy space and I’d have one enormous group of unrelated pages for future reading.
An obvious solution is to offer some kind of bridge between tabs and bookmarks, but I think that it might be more advantageous to make it feel more like the Read It Later add-on, a wrapper for Firefox’s bookmark system that allows users to easily create and maintain a chronological stack of unread items.
It would be great to have something like that, but with a more elaborate timeline view that would allow you to explore other browsing history that transpired around items that you saved for later reading. Similarly, it would be useful to be able to have tab groups "expire" and shift collectively into the reading list stack after a certain amount of idle time. This could have some kind of Weave sync capability so that users would be able to easily work through their reading list from a mobile phone.
Taken together, the underlying concepts behind Tab Candy, Tree Style Tabs, and Read It Later hold the potential to revolutionize Web browsing and solve a wide range of the tab management and information overload problems that are faced by users.
This post has been written by Ryan Paul on July 29, 2010 10:25 AM couresy of arstechnica.com.