OSNews

Thanks for the comment. Unfortunately, "advanced" is not a good idea for any UI. Havoc Pennington blogged about this explained it a few years ago. http://ometer.com/free-software-ui.html

Unfortunately, "advanced" is not a good idea for any UI.

But what about advanced users?

Yes, but I disagree wholeheartedly with Havoc ideas on UI usability. They are not Gospel, last time I checked. To me, preferences and advanced options are essential (that's probably why I use KDE).

So the question is perfectly legit.

Actually at the link you provide, Havoc does not say "advanced" is always bad. He says;

"Advanced tabs must not be used as band-aids or excuses for stupid designs"

But he also says;

"Sometimes there may be a good reason for an Advanced tab or the like "

I think an advanced tab is always essential, particularly for free software. The reason is that you need to segment your users into users who are just users, users who are contributers and users who are developers. Then you need to turn users who are just users into contributers and ultimately developers. To do this you can't hide the internals of your program away from your users. You have to expose a certain amount of the complexity to them. Users have to fiddle, tweak and customize to be drawn into bug fixing, contribution and development. This is not for all users, but the possibility has to be there. Plentiful preferences, just below the surface in an advanced section of the configuration programs is a way to provide this possibility.

The problem with many preferences is not generally the number of preferences, but their organization and exposure to the user. Hiding the advanced preferences is a separate program/tab is a good idea.

Then there is the issue of FOSS culture as opposed to proprietary software culture. In proprietary software there is a clear distinction between users and producers of software and users have no business meddling in the production process. Unless they are invited in, under carefully controlled conditions, by the research department. This leads to an as vs them mentality between users and developers. I think this mentality is alive and well in the Gnome community and it is developing really well. Too bad. In the long run this will not be good for Gnome.

Yes, the 3D Cube is awesome and think it should stay on by default. As long as switching has reasonable keystroke combinations, and the switching stays fast, I see no loss in productivity, but only an insanely cool new view.

Thing is, Beryl requires 50 additional MBs of RAM than a plain 2D X server. The fewer plugins, the better, in terms of RAM.

One thing I'd disagree with is the 3D cube should not be off by default, this effect is a symbol of the 3D desktop on the Linux box!! It's to Linux 3D desktop as the dock to Mac or the start menu to windows.

I dumped the cube for the new desktop wall plugin. Personally find it more useable seeing all desktops at once, and dragging windows among them. And it maintains a nicely conservative yet still slick bling factor that doesn't get dismissed as toyish when I use it at work.

The cube is just so, you know, 2006.

1) Agree strongly. Riding rough-shod over existing WMs' methods is a sure recipe for chaos, as I've seen with both beryl and compiz (which I actually use).

2) Substitute KControl and I agree with you

3) Substitute KWin ... well, lol @me but I prefer Metacity's themes, so I quite agree. Let's work *with* the WinDecos, not against them!