Slashdot Mirror
Linux Mint Developer Forks Gnome 3
An anonymous reader writes "Clement Lefebvre, the Linux Mint founder, has forked Gnome 3 and named it Cinnamon. Mint has experimented with extensions to Gnome in the latest release of their operating system, but in order to make the experience they are aiming for really work, they needed an actual fork. The goal of this fork is to use the improved Gnome 3 internals and put a more familiar Gnome 2 interface on it."
The goal of this fork is to use the improved Gnome 3 internals and put a more familiar Gnome 2 interface on it.
TFA actually says that it is a fork of the Gnome shell rather than the entirety of Gnome. Presumably, it would be built against and installed along with the official libraries and applications. Just a single component being replaced; a bit like changing the default browser to Firefox.
# cat
Damn, my RAM is full of llamas.
Unlike what the summary suggests, it's not a Gnome 3 fork but just a Gnome Shell fork. With the whole back end untouched, they should be able to keep compatibility issues to a minimum.
You've made a number of unsupported assertions there. And of course since you talk as if you know what you are talking about you've been modded up.
I can't disagree with your take on the politics. I do take issue with the technology. Gnome certainly has had problems with being over-designed and over-abstracted. And I won't argue with your assertion about stupid UI choices.
Compiling Gnome, though, is pretty easy using but time consuming using jhbuild. Most users of course aren't affected in the least by the build process. Qt's build process is self-contained, but takes hours still. The end result is really the same for end users. Having every widget toolkit re-implement every wheel is fairly tiresome. Why not use lower-level libraries like libxml that already work well, and most importantly, are C-based.
As for the language, basing it on C was a wise choice. It's a far more portable language than C++ or Objective C, and *way* easier to bind other languages too. The GObject model works very well in other languages. Programming GTK+ in C++ is a joy (doesn't need moc either). GTK+ in Python is slick too, and actually manages to be fairly pythonic, unlike PyQt, which is really just C++ code in a python syntax.
Writing new GObject code is a chore, since there's a lot of boilerplate code to implement vtables, etc, but using GObject apis in regular C code is quite easy. I don't think Gobject is a BS OO extension anymore than C++ is. Functionally and under the hood they are fairly equivalent. No language support is a pain, but Vala is nice for providing that. I basically consume GObject code in other languages, and there has never been any issue there.
The tl;dr version of this post is that when you say that Gnome has made every mistake possible and that C and Gobject are responsible for Gnome being in a sorry state strikes me as being a rather baseless claim.
There's no file system for your GUI, is there? You can't cat /proc/pid-6939/window-2/grid-3-2/textarea-2.
There's Gnome Virtual File System (gvfs), which IMHO is the second worst decision the Gnome people ever did (the Gnome Shell iPad fanboi UI being the worst).
When the superuser can't access all files on a system, something is worng. Backup programs and automated root "find" commands fail because of ~loggedinuser/.gvfs which they can't access. Good job. And no, it's not all the other well established tools that should change to accommodate gnome. It's gnome being stupid and breaking things.