Why do we use tabs?
On Windows and Linux (this argument applies much less to OS X), there is already an app that exists to switch between various windows — the taskbar/window-list at the edge of your screen. It is specifically designed for this purpose.
More is wonky about the tabs in many cases, though, and specifically Firefox. The visual desktop metaphor is weirded out by putting the buttons that affect the viewport above the tabs themselves. I’m not sure how to say it, but the generic URI field affects a specific tab, rather than the URI field belonging to a specific tab. This would be better if the tabs were above the main toolbar, to show that each button belonged to this specific tab, and affected only this tab.
Still, why not just use the app that is entirely designed for the purpose of managing different windows, and stay consistent with the stupid desktop metaphor?
I think I can explain it for Windows: there is no concept of different “Desktops.” It’s definitely annoying to have to go through a bunch of Firefox windows when there’s everything else mixed in there too, especially when there’s not much of an indication which you’re on in the Alt-tab thing. But think about a more featured system like GNOME or KDE: I can use the scrollwheel on my mouse to quickly go between all of them, or alt-tab, or alt-(shift)-escape — all without other windows with other apps in the way. I see the same title in each Window as I would on the tabs, and have more flexibility if I want to see them side-by-side.
I also gain more than 20 pixels on every page I view by not having a tab-bar; I haven’t used any more info, either, because I already had the taskbar app visible.
Of course, on OS X, this is all hogwash, because there’s no sense in maximizing, so it would just create more clutter. I guess I’ll wait ‘til 10.5 to comment on that.
