FirefoxMobile-SmartTabs

From Medien Wiki
Revision as of 11:56, 27 April 2010 by JanD (talk | contribs)

Smart Tabs for Mozilla Mobile

Project by Jan Dittrich

"How can the way we make the tabs better representing the pages?"
Goals: make tabs better representing the pages.
No-Goals: Change the general model of interaction with tabs, Use signficant more screen real estate.

Current Situation

MozillaMobile 1-1 Tabs.png

Ideas for Improvement

Put the closing-Button on the right Tab side

Relocate Button Wireframe
  • Pro: The more relevant left edge is free.
  • Contra:
    • The left side is closer to the edge which means the area can be touched easier since one does not care about tapping a close object on the left side.
    • When dragging the Bar in, one sees just the Buttons at first.


MozillaMobile pageCrop.png

Smarter Page Crop

Concept: ???

  • Pro: Coosing a more representative part of the page means better recognition and more differenciation between similar pages.
  • Contra: ---


MozillaMobile addFavicon.png

Adding Favicon

  • Pro: Visual grouping of same Domain-Pages. Recognize logos of commonly used webpages.
  • Contra: Could be mistaken for a button if not carefully integrated.

Note:It is likely that the favicon itself is hard to recall for non-frequently used pages. But it often gives color clues for the page. (Further research is needed)

Better rendering quality

MozillaMobile betterTabImageRendering.png
  • Pro: A better rendering of the tabs will improve the recognition.
  • Contra: Performance issues. Could possibly be solved by rendering low-Fi first and when e.g. the user reads the page, the hi-fi Tabs are renderend.

Retrive closed Tabs easier

  • Pro: Provide a Button oder Overlay-Button for retreiving an accidentally closed Tab.
  • Contra: Takes valuable screen real estate.

Wireframes

Study 1: What do people recall about webpages

Conducted by Bastian Bügler and Jan Dittrich

Goal

...is to find out what people can recall about websites they visited previously (about 2-10 Minutes ago)

Methods

Participants had do do simple tasks on several webpages. After a masking time of a couple of minutes filled with other tasks, they had to tell about the features that they link with the webpages.

Results