If you’re in Georgia and reading this, the title should already be known to you.
In the meantime, I’ve been failing operating systems (something which I hope to improve this week), but am improving in the other three classes. I guess that I’m more of a desktop kind of guy, especially since I’ve been reading up so much on the desktop Linux movement as is being moved for by Linspire, Canonical (Ubuntu) and Novell (SuSe).
However, as good as “desktop Linux” may seem to be, and as much as I’m studying about the nature of most distributions of GNU/Linux, I’m not really sure if I want to try it out again, let alone make a long time commitment to it.
Now, don’t take this as saying that desktop GNU/Linux doesn’t have any financial backing, or that it will have to make itself more appealing to Windows users in order to command a serious presence on the desktop. All of that has been debated time and time again, with both sides agreeing to disagree because of deep-seated prejudices, security concerns, and that all-important POSIX compatibility.
I see this one particular, glaring problem with desktop GNU/Linux at its present state: the desktop experience is still based upon the internal file system.
In fact, I also see this problem with Windows as well.
See, if there was one thing that Apple did right with Mac OS X, it was to separate the desktop and desktop apps from the file system. You really wouldn’t even know that OS X has a Unix base (known as Darwin) unless you took notice of the terminal, which doesn’t feature big on the OS X desktop.
Also, http://developer.apple.com/macosx/architecture/index.html
Plus, on the more complicated side, Apple chose to use Quartz, a graphics layer that is based upon, and utilizes, Adobe’s PDF format in order to generate windows on a graphics-model-agnostic level (unlike X11, which is based upon raster/bitmap images). This, they chose over X11, the standard windowing system for most GNU/Linux distributions, although X11 capability can be added on if its necessary (say, for GIMP or some other non-Mac-native app).
Finally, compared to Apple’s very-well-thought-out desktop (although the same can’t be said for the Unix underneath, due to the microkernel structure that is utilized by Darwin’s XNU kernel), I find applications for KDE or GNOME to be lacking, due to the first reason mentioned above. All applications for OS X that you see or hear about on a daily basis are based upon, and rooted into the GUI itself rather than the underlying Unix. In fact, the structure of the desktop (known as Aqua) is made in order to interact/come-into-contact as little with Darwin as possible, thus allowing for a self-contained desktop framework.
However, with GNU/Linux, you have to deal with package managers, or look into the repositories of your favorite distro, in order to see if KApp 1.5.x has been made available. Oh, and don’t forget those libs, or else you’ll have a broken app (if you’re lucky).
That, pretty much, is where I find fault with desktop GNU/Linux at its present state. I mean, maybe someone could “hack” the .ODF (OpenDocument format, native to OpenOffice.org) to create a portable document format that will be free (unlike Adobe’s PDF, which is proprietary) and openly-spec’ed (call it “.POD”), then build a windowing system (maybe extending X11 by a long shot) for .POD that will render graphics as well as Quartz/PDF does for Mac OS X.
Then, after that, create a self-contained desktop layer that will use modified QT + XUL + some-OOP-framework, and, using VLC/XGL/Xiph.org-audio-video-image at the core for media/graphics, build (1) a be-all-end-all library and application framework and (2) what I’d like to call “one great-whopping-big application” which will be more of a stratified runtime environment that will be, essentially, its own application management system (one that, mind you, will allow you to have/install/uninstall applications of any and all types without ever having to know about, or interact with, the underlying Unix or its own package management system, whatever it may be).
However, after all of that is said and done, I would think that a whole company that is dedicated to building decent desktops (and whatever could be built upon it, like the Web) would have to take on this matter, as I don’t think that the desktop GNU/Linux movement would ever push the envelope in the aforementioned fashions.
Oh well….
