I know this will sound crazy but with the whole 'windows' GUI environment that Microsoft/Macintosh introduced to us with all the boxes. 'nix developers are in a position to radically change to something more ingenius. I don't know what though. I don't understand why after 20 so years all OS GUI's look the same, yeah you can change colours and icons and where the menus are placed but it just doesn't seem to cut it. even Vista is just a variation.
going from feeding in punch cards to a dos text comand line to GUI's like Windows3.1... then what's the next step?
I know there's alot you can convey through a typical monitor without munching on resources.

I'm just worried that Windows year 2050 viewed on holographic dispays will still have us shutting down through a start button and arranging little windows in a fixed space.

I know i'm gonna get a lotta flak for this post.

Just throwin' it out there.

Regards, Screphoe

The shell command line rules.

Just sayin'.

People do try other things.
have you seen any of the tiled window managers?

here's a couple:
wmii/dwm
ratposion

Problem is, as always, it's what people are used to.
People struggle enough with the new M$ office menu.

I sit at work on a CAD package all day, so for me, windows mean nothing...2 virtual desktops are enough(one for Microstation and one for Firefox).but of course I'll have to wait for Vista for windows to provide that.

I have an idea!
Create a so called RTS-desktop.

The idea is basically to have a desktop similar to classic RTS maps like Starcraft or Warcraft (or Supreme Commander :P).
[Yes, I know those are all proprietary games for Windows, but I haven't completed my transition to Linux yet, and those are some of the best RTS games there are.]

Here are some of the things that could be implemented:
-Huge desktop, screen only displays a part of it and you can move around by placing the mouse at the borders.
-Zoom in zoom out (like in XGL)
-Minimap where you can click a region of the desktop directly.
-Icons on the desktop can be moved around like units in RTS-games (useless, but fun and can be used for some kind of game)(M$ icons vs FOSS icons for example ^^)
The icons could be given some properties like speed, power, life, etc.
-Possibility to select windows of the same type directly or to make groups of windows
-Possibility to give orders to individual windows or groups of windows to go to a certain point of the desktop
-Window formations
-Window with several sub-windows
-Physics engine for the windows with configurable parameters
-Multi-user desktop through the network, where the windows would have different skins depending on the person using them.
-Several users can "play" on the same Desktop over a network
-Nuclear launch to close all the windows at once ^^
-Teleportation gateways through which the windows could be teleported from one place to another
-Intelligent windows gaining experience points and placing themselves better and better on the desktop by analyzing the user's habits.
-Possibility to destroy windows with other windows

I don't know how useful these implementations can be, but as an RTS fan, I would immediately get a Desktop system like this.

Some ideas may be a little bit crazy, but the huge continuous desktop with minimap(or minidesktop) might be a great change to the classical separate desktops currently used.
And the multi-user desktop where each user has a personal skin for the windows he's currently using might have practical applications.

Another possibility would logically be an "FPS-desktop".
It could either be a simple 3D-desktop where you move around using classic FPS controls or a 3D game like Doom 3 with readable and usable PC screens.

But I am especially waiting for the hardware user-interfaces to change.

Things like the Wiimote for the PC (video), the Optimus keyboard, "touch-tables" like in the movie "The Island" (unfortunately full of M$ advertising), the cool glove interface from "Minority Report", 3D displays (where XGL will become even cooler), etc...

And maybe someday a virtual reality system which is not too expensive.

P.S:
What is the best RTS on Linux?

I know this will sound crazy but with the whole 'windows' GUI environment that Microsoft/Macintosh introduced to us with all the boxes.
AFAIK it was Xerox that gave us WIMP (windows, icons, menus, pointers).


'nix developers are in a position to radically change to something more ingenius.
IMHO the real question is: who will need or benefit from "radical", or even worse: "ingenius", changes? If you're familiar with the history of art (as in expressing something), design (fitting functionality or "less is more") and ergonomy (benefitting human comfort) you see there is a big difference between things that are "radical" in one way (say Raymond Loewy's train designs, Brancusi's Bird in Space or Le Corbusiers chairs) and things that are "radical" another way (say the old Alien / Blackbox Photoshop plugin UI's, Albert Speer's design plans for reshaping Berlin or them cubic whisky glasses). Things that are radical in a harmonious way appeal to us because they challenge or invite us to see them as extensions of what we are: think of how something with the form of a pebble fits snugly in the palm of your hand. Things that are radical the opposite way are hostile because they are not adjusted to how people think or act (confusing) or say out of scale (threathening).


I don't know what though.
Fogetting for the sake of discussion people can be trained and conditioned to work with things like WIMP, what in your experience (in the widest sense of the word) would be a good, new and unused metaphor for locating an item with certain attributes on a remote system, accessing it and then moving it? Got one? OK. Now imagine you, for that sequence of actions, make a macro: how would you access it so it's top of the list of macro's? How many formal steps would it take you to explain that to me over the phone? Do you need to explain your metaphor first or can I relate to it immediately? If it did you should seek a carreer in interface design because clearly you're a natural. If it didn't it sort of illustrates the problems that come with making things fit the human way, *why* they have to fit (and not the other way around) and so why you so rarely see "radical" designs that benefit and inspire us.

Moved: This thread is more suitable in Linux - General and has been moved accordingly to help your thread/question get the exposure it deserves.

--jeremy

3D VR desktop. Something like booting straight into Second Life:
http://secondlife.com/community/linux-alpha.php
Say you want to control your computer radio, you could just hop in your VR car and use a real-world 3D interface you're already familiar with. Talk about a short learning curve! Want to share a song with your buddy over the internet? Just call him on your VR phone (VoIP) and meet him at his VR house (streaming audio).

The "Sugar" interface designed by Red Hat and MIT for the (Linux-based) One Laptop Per Child devices is basically a rethink of how computer graphical interfaces should work, and may be very influential in time. Some or all of it could probably be reused on Linux systems for standard PCs if somebody wanted to do it.

Another interesting project is Sun's Project Looking Glass, an attempt to devise interfaces that integrate 3D in an effective way.

Less radically, various people are tinkering with taking the GNOME desktop in different directions to improve usability beyond what the conventional Windows/Mac-style interfaces offer. KDE 4 may also include some nice refinements as well.

Having said that, I kind of agree with unSpawn. Familiarity sometimes beats innovation - the QWERTY keyboard layout is still the standard despite Dvorak etc., and I suspect that windows and icons will also be around for a long time even after better hardware is devised for human-computer interaction than the mouse.

certain DEs and WMs are trying to break the mold. Have you checked into mezzo at all?

But it does really come down to how people are conditioned. We use them because we are used to them. Differences often breed contempt. People get frustrated when they don't know how to use something (something I felt when I first tried ratpoison). They'll go back to what they know.

It's hard to break a computing modality which has been around for as long as it has. Do I think the current system is the most efficient/best? No. Are the subtle variations we get with the development of QT4 and the like and the widget systems a step in the right direction? Of course.

After all, any major revolution in any system has a large chance of reject. Big systems tend to change a bit slowly. Also, I might add the old saying, "if it ain't broke, don't fix it." People have been reading for a great many years. While language has changed (languages and lexicon), the basic concept of a "book" tends to be the same: bound paper with little symbols on it. I have a feeling we're going to be stuck with the fixed space and little windows for a long time, but I do wonder what will be after this. 3-D interfaces... mind control... who knows.. Maybe I should go ask Big Brother Bill.

The interaction between user and machine has made new users not realise how computers work.
Big graphic intensive GUI's are a waste on resources and user productivity. Computers in my mind are just file-cabinets that have alot of tricks up it's sleeve.
The interface should not draw attention away from it's primary tasks.

3d virtual desktops like second life or the RTS one are a total drain on valuable resources and user productivity, i mean i don't want to have to run around in a vr world to do something that should be a few key presses away.

And they can shove vista somewhere else, because i can't wait 5 minutes for an os to boot up just to send a quick e-mail

With "sugar" for the 'One laptop for every child' poor children in africa need farming equipment and not laptops. Even so an os like 'sugar' will only further confuse them when it gets down to basics.

OS's should'nt dumb down users control further than they already are.