It's not a OS by itself. You have to have other components... like a kernel, init, shell environment, C language libs, etc etc.And when you add all those together what do you have? I'll tell you... its called a UI. Not an OS, not by ANY stretch of the imagination.
But it's certainly more of a OS then the Linux kernel is by itself.
By the time you get to telinit 3 your already using a massive amount of software that is completely outside of the Linux kernel.Lets put this into perspective.... If you start in runlevel 3, does the computer still process? If the answer is YES, then GNOME is NOT part of the OS, but rather just junk added on top of it.... specifically, the UI.
The part that is 'linux' is that 30MB or so vmlinuz file in your /boot directory, the modules inside of /lib/modules and a few other odds and ends. Meanwhile the vast majority of code your running on your system, at this point, is actually written by GNU.
That's why people made a big deal of 'GNU/Linux'.... because without the GNU part you can't actually build or run any software, nor do you have a user interface.
Don't forget we are talking about _DESKTOP_ operating systems. Not just something you can run OpenSSH from.
This the entire point of the discussion. Should Gnome just support Linux? Should it still support swapping out parts underneath it?
If that is true then should Gnome try to aim to work like the Linux kernel were the developers are free to improve and specific versions of the software for the rest of the OS in order to get best performance and best user experience?
In particular the Init...
The current approach we take with Linux with distributions and such provides a decent enough OS, but it's not good enough.
A couple of examples:
The Linux Gnome/KDE desktop pre-dates OS X, but OS X was able to achieve things within one or two releases that Linux still is struggling with.
Meanwhile we have Android which within a couple years have a huge number of software packages and has very good commercial success in mobile products. The Linux desktop has not had anything remotely close to that sort of success despite being around for decades.
When Linux mobile developers tried to adopt the traditional Linux approach to mobile OS... it flopped. I think the ideas and approach for things like Meego is pretty cool. Lots of cool things and probably more open then Android, and it fits better with my personal aesthetics to a OS.... but you have to realize that this is not a new development. People have been working on turning out a mobile Linux distro for years.
Nokia has had commercial products out there using Maemo since 2007. Development started in 2005 or earlier. There was efforts before that.
My point here is that I think there is distinct merit it just trying out a different approach. Just concentrate on making a Linux-Gnome desktop the best desktop it can possibly be and not worry about anything else besides that. Don't worry about how well it runs MySQL or runs apache, don't worry about whether or not the software is portable or Debian/Redhat friendly or anything.
Just create a intense focus and work towards a specific goal instead of trying to be everything for everybody.
GNOME in 1999
NeXTStep was released in 1993
So NeXTStep was around about 3 years before KDE came about. MacOS X heavily is based on NeXTStep/OpenStep.
So to state it "pre-dated" OS X is kinda a lie. Yes if you are treating OS X 10.0 as a new desktop it's younger than KDE/GNOME (being released in 2001), but the underlying API is MUCH MUCH older.
The same thing can be said of Gnome or KDE.
Very stupid move indeed. There is BSD, and other platforms. Even windows. Even if I hate it and its company.