> Now, I would really like to hear the *reasoning* against having > default categories. Why it is preferable to let the burden to the user > do all the job for each new applications he/she installs (and all the > installed applications that come by default). Sure, it is impossible > to find a good categorization that fits everyone. But we can find a > reasonable initial one. The current one is good enough for me.
Display a flat list of all application launchers, excluding system tools and file viewers. Here's my reasoning: The default applications list should exclude system tools (because they aren't applications from an end-user perspective) and file viewers (because they aren't useful when launched directly). Ubuntu, then, comes OOTB with fourteen apps that belong in the launcher -- aisleriot, calculator, gedit, brasero, empathy[1], firefox, gwibber[1], libreoffice, mahjongg, rhythmbox, shotwell, simplescan, thunderbird and transmission. Using the XDG categorization scheme, those applications get divided into six categories; some categories have only one application. The small number of items per category significantly diminishes their usefulness, and there's no sensible way of consolidating them. Categories are also much harder to pick out of a list than applications. Applications have recognizable, definitive branding. Categories lack such branding: the spinning fox is always firefox, while a globe could mean many things besides "Internet". For user-installed applications, the branding is extremely prominent in the software center, so we can expect users to recognize it. The dash can display 24 icons per page at it's smallest. On a 1080p monitor it can display up to 84 items. That's plenty of room to fit the default apps plus several more installed from the software center. As long as a flat list of all applications fits comfortably on the screen, a user can pick an application out of the list without needing to interact with the computer at all [2]. [1] The presence of empathy and gwibber on this list is questionable since they come with system branding (the "chat" and "broadcast" features of the messaging menu) rather than application branding. [2] Reducing the amount of input is more important than you might think. Every output->input cycle requires an inexperienced user to shift his attention from the screen to the mouse and back. If the entire list is already on the screen, choosing one is a simple task. If he must point at a category and then choose an application, it involves multiple subtasks. -- Mailing list: https://launchpad.net/~unity-design Post to : unity-design@lists.launchpad.net Unsubscribe : https://launchpad.net/~unity-design More help : https://help.launchpad.net/ListHelp
