If things don’t work out talking changes within, then you can always fork it.

Good luck.

It works just fine and I don’t have to worry too much about viruses and the like. Open Office is much like Microsoft and I love the wobbly windows and special effects.

Go Linux !

Invite artists, graphic designers, user interface architects, psychologists, workflow analysts, 15 year olds, etc… they should brainstorm and come up with sexy, impractical, wonky visions of new form factors and ways to work with software.

The programmers, system designers, software architects should ignore what we have right now and design from the ground up new toolkits to construct said ideas in a modular and robust fashion (CSS so artists can build the look – who could have thunk it!) as well as re-inforce best practices (testing, stability, refactorability)

Then with that vision of some utopia – then with the project managers and architects build a road map that implements the best 20% of the toolkits into Gnome and let people at it.

There is no reason to de-stabilize the desktop…. but this resistance to change from old schoolers has really made Gnome a boring place to be — and I don’t know about you but I hate working in a grey cubical no matter how stable and predictable it is.

It would seem there is so much negativity out there towards the creative and slash and burn types that by the time the complainers catch up with the visionaries we will be 15-20 years along and still 3 years out from a new desktop because no one has created a vision…

There is much improvement needed to the desktop – the 1-2 inches of wasted space with status bars, toolbars, menus, etc is indicative of developers and users spoiled with 20″+ screens. The programs, while functional, are quite bland and un-inspirational (do you want to work in a grey cubical all your life???).

Those trying to fit into other form factors, or leverage other bling like 3d, CSS, and animations will refine the desktop in ways not quite visible to those in plaid pants panicking over their desktop being taken over and pimped out leaving nothing working in the wake.

Sure – the desktop metaphor hasn’t changed – but you would be hard pressed to find anyone wanting to go back to 10 years ago. I’m not even sure I want to stick with a platform that doesn’t show interest in improving itself.

Not going to happen unless novell want to kill of is linux division. QT 4 is or will be gpl 3 soon, the fsf made some last year provisions to gpl 3 to punish novell for their linux sell out to microsoft. If gnome switches to qt then the fsf will be able to revoke novells license if they publish a new linux distro based on that qt gnome..

first I want to thank you for your work on Jokosher. Splendid piece of software.

I have been using gnome from the beginning, and I have to say that the last couple of years it really starts to be usable. It looks quite nice, simple and while there are still significant improvements being made under the hood (GVFS), it seems to basically done. I haven’t used windows for over ten years now, linux and gnome just do all I need.

People complain about Vista that it is slow and not very groundbreaking (probably true). People complain about lack of vision in Gnome. People rave about how wonderful OS X is, esp. its Aqua/Cacao interface and the dock.

I think that all three systems are “done” in a way. There is not much more to do with the WIMP interface. Sure, OS X is somewhat 3D and has a really nice programming interface and toolkit. Vista is also 3D but it does not really add anything to usability. Linux has Compiz, which at least is useful for multiple desktops and finding your windows.

But when the cube stops spinning, you still just have a flat desktop interface. Smooth scaling SVG icons, anti-aliased text, and Avant Window Navigator adds a nice OS X dock look-alike including ’stacks’, but still we are stuck in 2 dimensions.

If you want to make bold steps into the future of GUI, two things obviously need to be incorporated:

1. Real use of the third dimension.

* in a non-confusing way
* integrates 2D windows in a natural and useful way
* onset of development of a 3D widget set, to make 3D applications
* no more icons; use scaled (or distant) windows instead. Just zoom in to work with, or let come to the foreground.
* you can group windows in space; like wise group files in a
 (floating) grid in space. This way you can represent directories by
  multiple levels of scale.
* no more taskbar, dock and such, but an object that appears by
 tapping screen, or some hotkey combination (say ESC-ESC). This (3D)
  object would give you access to your apps and menus, and might
   look a bit like a "Pie menu" (see wikipedia). I always liked the
    right-click menu in windowmaker.

1. Touch and Multi-touch interface

* most people have no touch screen, let alone multitouch, so normal
 mice should remain usable for quite some time.
* there are different types of multitouch, including a type (from
 Sharp) where each pixel is also a sensor. Imagine logging in by
  simply placing your hand on the screen, reading your palm-print?
* use of a number of standard gestures for different actions. (for
 instance "pinch" to zoom, "swipe" to move through space or group
  files, etc.)

But for something really wild, let me introduce me my idea of “The Cat” (no longer very relevant to Gnome, more to user interfaces in general).

The Cat would be an avatar of an AI that is actually an expert system. The avatar generally walks around a bit, stays in the distance and ignores you, sometimes disappears, just like a real cat would in fact.

The expert system behind it is somewhat like wikipedia, in the sense that it is tought by the people using it. What it does is help you with computer related tasks. Some such tasks may be: - replacing a failed RAID1 device. - making a Latex document - organizing your directories and files - color correcting you display and scanner with LProf and other tools (if you are a graphical artist) - writing your first C program - advanced programming using subversion, profilers, debuggers etc. - doing a mail-merge.

The expert system would have modules for each of those areas of knowledge that are added by people in that field. The system also learns, and uploads this knowledge to the (somewhat wikipedia-alike) backend. The hard parts are: 1) codifying knowledge in some way an expert system understands (but that’s already solved in all kinds of existing expert systems). 2) a method that lets the Cat interact with programs (like filling in values, checking boxes, clicking buttons). Also this ties in with (1). In windows at least there are ways to automated program configuration and installation by replaying. Now replaying is way to simple for the Cat, but must be a part of it. 3) Implementing the “learn mode” of the Cat. If it is learning it must observe the user and record his actions while the user describes them. 4) The cat must tie in to the system itself; i.e. knows about processes and files, can recognize out-of-disk and out-of-memory situations, can deal with a run-away firefox process for instance, or can warn the user when a hardware fault happens. 5) Not making another Clippy. The Cat should never annoy the user. It just keeps silent company and even when there are real problems stays discreet and never takes over control of the machine. The expert system may present a solution but the user is free to alter it, or ignore it altogether.

As time goes by, the Cat should get more and more input from users around the world, teaching it about all possible uses of computers. It might be made smarter over time, adding the ability of limited reasoning and recognizing generalized situations from examples.

People who don’t like cats are SOL. Or maybe a snake could be implemented for them.