This is one of the problems that faces the modern desktop computer that needs to be approached.
First of all, let me state this: Tabs are a list of different documents being displayed in a single window. It is a form of navigation within an application. People are becoming used to tabs. This makes list views that can be more useful less learnable, as people have already learned how to use tabs. It's something called familiarity. When a user makes the switch from a windowed-system like Microsoft's older versions of Internet Explorer, the users wanted functionality that was familiar to them. This included having a close button on every tab (as they had had on every window). This not only makes the functionality much more visible, which I find is the MOST IMPORTANT aspect of usability (keep this drilled into your heads please), I don't have to move my mouse around so much to close a tab. I also don't have a middle mouse button and find clicking both mouse buttons simultaneously to be awkward (once again, familiarity). Still, considering there are differences that users have with the tabbed interface, it is important to let it be flexible. This means the user should be able to modify the appearance and behavior of a tabbed interface. Not only does this make the interface more flexible, it gives the users more control over their environment. While it is obvious that the GNOME community wouldn't want to bombard the user with options for everything in one screen, this can be controlled by grouping functions in a way that makes sense, not to lack possible future innovation in the ways they are displayed. Being able to recover from doing something stupid is another usability must-have. This means you should be able to easily recover from accidentally hitting that close button on the tab. Consistency is another important feature of usability. In this sense, tabs are out because there are so many different implementations of them. What can be done about this? Let's see... how about the GNOME developers, KDE developers, and Firefox developers work together? Don't forget to include everyone else that employs a tabbed interface. Still, everyone seems to be interested in tabs and that's where the most usability advancements have been occurring. None of these improvements have been occurring with any other type of interface that can be used to accomplish the same tasks that a tab accomplishes. I would like to see some improvement in that area. Just because some functionality isn't right for some people doesn't mean that it doesn't work for a lot of people. Having a functionality that can be useful and not using it is better than not having a functionality at all. For instance, I'm building a LEGO MindStorms NXT robot for my HCI class with someone. We're designing the interface top-down since we can't build the final version of the robot yet. In our top-down design, we are including all the functionality we might want to use even though it might not be used in the final version. Still, we see that there could be some use for it in approaching the problem of usability. Different people could build a robot that performed the same task using some of the same base functionality that we created but didn't end up using. That's pretty much the reason user settings were created. We have this, and we find that it could be useful, so use it if you want to. When you stray more from the kernel and into the higher levels the flexibility/control concepts are less-utilized. At the level of a window manager, GNOME should recognize that there is still a lot of flexibility and control that can be implemented. An example of where a tab's usability-improvements could be useful, but not necessary: I would like to be able to put min/max/close buttons on my window list buttons. No desktop currently makes this directly available to me, meaning I'm probably going to write it myself sometime in the near future. The same concepts that are used in tabs could be applied to any kind of lists that can be used for the same purposes. This would also help the familiarity level in a user that's using the interface. In other forms of consistency, it would be good to have a hierarchical key-assignment manager. Something that could tell a program "Oh shit, that's being used by on a lower level than you, inform the user and ask for a different key!" I find key-assignment managers are quite limited these days. I also hope someone copies Windows Vista's hierarchical volume controls, although I know it will probably be done in the next year. On Tuesday 27 March 2007 04:57:24 pm Erik Jan Philippo wrote: > Hey, > > It's very annoying that all the tab based application behave different. I > think we should create a standard for tab behavior at freedesktop. The > standard defines mouse and keyboard behavior. Several open source project > should collaborate with each other. I suggest mozilla, kde, gnome and > novells betterdesktop. The standard should be tested with usability tests > (that's why i think betterdesktop.org is usefull). If the standard is > defined we should ask as much as possible open source projects (if they use > tabs) to implement it. Off course we provide them with the results of the > usability research to convince them. If kde, gnome and mozilla agree with > it, probable a lot of small programs will also use it. I hope this will > lead to a very consistent tab behavior in every open source applications. > Before I'm going to spend my time with it, I like to know what you think of > it? > > Ubuntu is also planning tab consistency: > https://blueprints.launchpad.net/ubuntu/+spec/tab-consistency > > There are some problems with this. Terminal use ctrl-shft-t while a lot of > tab based programs use ctrl-t. The problem is that nano also use ctrl-t. I > think there will be more complications like this. > > Greetings, > > Erik Jan > > > _______________________________________________ > Usability mailing list > [email protected] > http://mail.gnome.org/mailman/listinfo/usability _______________________________________________ Usability mailing list [email protected] http://mail.gnome.org/mailman/listinfo/usability
