The future of Gnome?

[s335328 student uq edu au]

Hi everyone

I've only been watching this list for a short period of time, but I'd like to ask some questions. Admittedly some of them may seem a little idiotic to a lot of people, and I can accept that. I probably should apologise up front for wasting everybody's time if it turns out that I've done so.

I'm both a Linux and Gnome novice. I've only played with Linux a little in the context of my work with Oracle, and only really know Gnome from the work I do within Linux when I use StarOffice. So technically I'm not really up to scratch, something which I hope to remedy as time goes on. However, I am everyday at work exposed to a heap of ideas about the future of computing (admittedly in large part from the way Oracle sees it). 

From what I can tell, both Linux and Gnome seem to approach things very much from a fat-client perspective. Install heaps and heaps of software in order to do the simplest of things. Now if I'm some guy on a warehouse floor who only uses an inventory application or a personal assistant who only uses a word processor, email and a personal organiser, what do I need with httpd, etc within linux and then this tremendously advanced desktop with ORBs and the like in order to do my work? 

Would it make sense to approach "the desktop" from a new perspective? One where the structure of applications is not tied to their presentation within a Gnome client running on Linux. I'll try to explain what I mean by this. What if I have a PDA, or a Palm V or Windows CE or whatever, where the desktop is inherently different to the one presented on my PC? Would it make sense for a "Gnome Server" to present content to me based on the computing device which I use to see it? 

Again, would it make sense for a HTML-ised version of the desktop to be presented to someone who connects to this "Gnome Server" with something like Netscape or the like? Straight away we have solved these two people's problems. Rather than the world's largest program sitting on their computer so that they can use the world's simplest applications, they can use whatever client they see fit, and all they need is a browser. They connect to Gnome and it figures out how to interpret the applications display requests in a way which fits with the way that application is being viewed.

A recent example of this type of thing is displayed in an article currently on technet.oracle.com. A company called Indus represents its GUIs with XML documents, and then they write interpreters for each client so that the GUI can be represented in a way which befits the particular client. 

Such an ideal leads to all sorts of interesting possibilities. Immediately within a framework such as this would come the ability to "share desktops", something which you need third-party software to do now. The complexity of the clients would be decreased by orders of magnitude as we install everything on the server and let whatever clients access it in whichever way they want. Fridges could display Gnome desktops as they begin to plug into the network. And the list could go on and on.

Basically, I know that all of this is very X-termish, but I feel that the internet has changed the way these things can be achieved since those days.

Basically, after all of this, the question I'm asking is...... 

Why does the desktop have to be so inherently tied to the way Microsoft has done things for so long? 

Cheers,
Sean