Planning to deprecate libgsystem, introduce libgsystem2 as git submodule

[Colin Walters <walters verbum org>]

I've updated this page:

https://wiki.gnome.org/Projects/LibGSystem

libgsystem has already become Linux specific (reading symlink xattrs from /proc), and I'd like to make it 
even more so (some container APIs).  

gnome-desktop used to depend on it, but with this bug:
https://bugzilla.gnome.org/show_bug.cgi?id=708453
doesn't any more.

Further, GLib now has the cleanup macros:
https://bugzilla.gnome.org/show_bug.cgi?id=743640
As well as GSubprocess.

Another major chunk of libgsystem is the file utilities.  However, I've
been backing away from using GFile * in my code, which operates
quite a lot on files.  The two primary reasons are:
 - Doing atomic operations and malloc() just to interact with the
   filesystem sucks
 - Inability to make use of openat() and friends, which provide speed
   and security benefits

So my high level plan is to make a "libgsystem2" which is
explicitly Linux specific and drops most of the GFile* stuff,
and it will return to being a git submodule - shared among consumers
like ostree/rpm-ostree, NetworkManager, xdg-app, libhif/PackageKit,
and anything else that uses glib and is Linux specific.

There are a few other things like GSConsole in libgsystem that I
never really pushed forward into GLib, but that API didn't really
work out anyways.

Speak now if you have any thoughts, objections, or just want to see your name in the mailing list archives.