Microsoft: Linux on desktops a threat

[hm_murdock]

just a thought. I'm not knocking Linux at all, but I think something that would really help it on the desktop is if someone were to develop a user friendly UI layer the way Apple did over Darwin.

Instead of using X11, use a built-from-scratch, modern graphics framework (display Ghostscript? an open source Quartz?), develop a Linux-based OS that's as full-featured as Mac OS X, using KDE or Gnome as the UI foundations.

For it to be accepted on the desktop, much of what makes it UNIX will have to be hidden to the user. They don't know, don't want to know, and just don't care about anything in /etc, or /dev. they just want things to work. The concept of "mounting" drives will have to be an automatic process. You put the disk in, and it's mounted for you.

They're making huge strides in usability, but unfortunately, not enough. Remember, that for it to make it on the desktop, it can't be "good enough" or "as good"... it has to be BETTER than Windows. Better in every respect, not just licensing and stability. Usability has to be ahead. The Linux world will have to break down and accept processor-wasting eye-candy, because users are accustomed to it now. It's what they want!

Make a GNU implementation of the Apple/NeXT packages so that a program and all its support files are contained in their own tree, and represented by one icon. Users can't be bothered with worrying about library dependencies and what folder a file went to. It should all be contained in one icon. This is easier than Windows. Drag one icon to install the thing. No extra files, no registry, no endless installers. Just boom!

I have no doubts Linux will kick MS's ass on the desktop... but they've gotta do more than just make it look like Windows!

[xyle_one]

i second that jimmyJames.

[DC]

quote:Originally posted by void main:
There's only one word that popped into my mind upon reading that quote:

DUH!

What a stupid quote. I mean that's like saying "I am afraid if I put a few ice cubes into this full glass of water some of the water is going to spill out."

Even more stupid. Something like "hey, if someone takes some of my marbles, I might loose some!".

JJ, for your plan to work you must either make the enviroment X-compatible (with all problems that come with it) or develop a port for pretty much all package. Both will be huge undertakings, not mentioning the development itself.

Besides, X11 needs to be improved and expanded (it's modular. I love modules. expanding the specs is so easy). X is in itself not a bad design, it's just that the implementation isn't quite perfect. But the server-client thing is essentially a good thing. It's great that I can run some program from home (= faster computer) and display the stuff on the computer @ school (actual physical location).

[mobrien_12]

quote:Originally posted by Jimmy James: Mac Commando:
.

Instead of using X11, use a built-from-scratch, modern graphics framework (display Ghostscript? an open source Quartz?),

XFree86 already supports display postscript.

[emh]

quote:Originally posted by Jimmy James: Mac Commando:

The concept of "mounting" drives will have to be an automatic process. You put the disk in, and it's mounted for you.

This is exactly how drives are mounted in Mandrake Linux now.

[voidmain]

quote:Originally posted by emh:


This is exactly how drives are mounted in Mandrake Linux now.

It is in RedHat too, but it's one of the very first things I turn off because I can't stand it. I'm old school I guess.

[choasforages]

does it work with usb, usb 2.0, and firewire drives, or have they gotten around to that kind of automagical mounting

[dot.this]

I don't know if they use X11, but Lycoris sure looks nice, and user-friendly. According to their web site, Wal-mart is now selling desktop PCs with Lycoris installed, starting at $200.00.

I don't know if it'll be a big hit with consumers, but we can hope...

[choasforages]

of course it uses X11, every other alternitave is in the beta stages, and a better idea of mine,would be to write  a gtk-qt. basiclly a mapiing of gtk funciton to qt ones. so that the ui has a more consistent feel/*and qt's menus are better then gtk's will ever be,  and interface similarty*/ im jusst wondering, what kind of hurtles would be faced by such a project, and what potential show-stoppers could present themselves