On 5/25/05, Bosshard Raphael (bosshrap) <[EMAIL PROTECTED]> wrote: > Topaz thoughts.. > > Again a silly questions to the gods of usefull interfaces; > > Is the desktop useful (or is it just a relic from ancient times?)
It is a relic from ancient times, but I also use it to access all my files, so it might be useful... > Most people I know use the desktop (the thing in the background where all > these funny icons are) to store data. They create folders with names like > "Music" and "Private" and "None of you business" and put stuff into them. Yes, often I also put files directly on my desktop. Those usually are the "I don't really want to keep them around, but need to do something with them now"-files. It is important that these files are directly in front of me, both because I want to access them quickly and because I want to make sure I put them away once I don't need them anymore. > But whenever they want to access them, they use the file-picker, because the > application they are using is concealing those nice folders. The desktop is > the least accessible place on the screen. Using the file picker does indeed render the whole idea useless (I rarely use it), so we should consider why people do this and how to improve this. One reason is application-based design. The idea of the desktop-paradigm is, that you are directly working with objects. Application-based design breaks this, because it forces another abstraction layer between the user and the desktop. The best example would be an IDE, which basically represents another workspace inside your workspace. It is not, in any way, _part_ of your desktop. Neither is gedit, it's just another application object on your desktop, it's not a part of the desktop. Our actual desktop so far basically consists of the following programs: - The panel (as some kind of shelf) - Nautilus (as the actual desktop and for managing folders) - The various invisible helper programs Before spatial Nautilus, it was even less. As you can see, there is nothing to actually deal with documents! My vision is, that we improve the usefulness of the desktop itself. Make it a real integrated workspace (like an IDE for users) with documents being first class objects. It is important to note, that first class objects and applications are not an either/or decision. Just like spatial folders and the file browser can live besides each other, you could still have an application like gedit as its own object, even if text documents on your desktop would open as individual objects (by default). Applications become just another kind of object, which you can manage with your desktop. Of course another reason is that the desktop is usually covered by windows. This is more trivial to solve with modern computer capabilities and it's somewhat of a shame, that we didn't do it already... Expose might be the best solution already (I don't know how it works exactly). Personally I don't use the show desktop button, because I hate manually un-minimizing all the windows. Instead, I simply switch to an empty workspace... > What about a different to store data? What about a "sidebar" (instinct > anti-longhorn reaction expected) with all folders in the home directory? > Plus, maybe, some "persistent search" folders (All images, all music-files by > Jethro Tull, conversations with ALICE...). One of the downsides of the desktop location is its limited space (hooray for zooming interfaces), with a sidebar you'd make this problem worse. I believe that ideally panels would behave like the desktop, so you could easily make your own shelf for the things you access the most often. If we'd have this, we could even think about making the desktop truly spatial for each workspace. (Metadata) queries will also make the desktop metaphor more powerful (instead of replacing it), by making the act of "cleaning up" less important (we could even automate it). It is certainly true that the current implementations of the desktop have serious issues, but you should not forget that it hasen't changed much during many many years. It does not mean that the idea itself is bad. Daniel _______________________________________________ Usability mailing list [email protected] http://mail.gnome.org/mailman/listinfo/usability
