Image of the glider from the Game of Life by John Conway
Skip to content

Gnome Integration

First, I want to thank all the traffic from the Gnome desktop devel mailing list. I normally sit around 20-30 hits per day. Your traffic has doubled that. While I am certainly NOT looking for traffic, it is somewhat pleasant to know that I am reaching out, and creating interest in someone. So, with that, thanks.


A recent discussion with Tuxgirl in IRC has prompted this post. The big question, is I keep eluding to tighter Gnome integration. What exactly do I mean and how would it be implemented? Well, let's see if I can shed some light on this topic.

To begin, I am not an computer graphics artist, so I won't be able to give you any screenshots or prototypes that are in my head. I'll just have to be extremely detailed in my post to convey what I am trying to imagine. Next, I want to get this out there, I AM a Gnome user. I prefer Fluxbox as my absolute #1 window manager to all others, because of sheer speed. Bug, Fluxbox is limited, and it gets on my nerves from time to time. My next choice would be KDE, because of it's direction, speed, and implementations. And the fact that KDE on Ubuntu is far less bloated than Gnome on Ubuntu and less bloated than KDE on other Linux dsitributions. My third choice would finally be Gnome. In a nutshell, here would be the breakdown of usage by percentage- roughly give or take:

Fluxbox- 40%
KDE- 35%
Gnome- 25%

As you can see, we are talking narrow margins. Gnome is still used heavily. Heck, I am in Gnome right now typing this post (mainly, just so I can be descriptive in this post). The fact of the matter is, I am a Gnome user, and I do enjoy using Gnome. So let's see if I can shed some light on what I think would be a tightly integrated Gnome desktop.


Although I use Gnome, I am definitely no expert. I am not aware of the internal workings of much of the GDM or the window manager. As such, if a feature that I request here is already implemented, please accept my ignorance. Also, I realize that many of these requests could cause security issues and performance issues. I do think, though, that those two requirements can be met efficiently and effectively. Also, just because a request here may utilize two or three programs already implemented, it doesn't mean that those programs need to be obselete. The greatest thing about Linux and Open Source Software is the freedom of choice. No one using the Gnome desktop will be required or mandated to run this software, it will just be an option that is already available. Lastly, I will be using KDE in almost every one of my examples, as well as proprietary software from Windows, because I think much of the softare implemented in those cases just works.


Instant Messaging: I pick this as my first topic, as it seems to be the most prevailent to me. I would like to see instant messaging within the intranet built right into Nautilus. There could be a Gaim-like interface as a sidebar in the window itself or as a seperate GUI application residing in Gnome. Of course, this would be turned off by default, to prevent the issue of opening ports on the network, and thus posing a security threat. Also, by turning off as default, we are keeping the processes minimal. However, firing it up would identify other Nautilus users on the intranet, and instant messaging would become very viable, even without manually adding users. This would also extend project management further and make it easier to manage data flow leaving the network.

Voice Over IP: Extending this IM capability further, installing Asterisk and using VOIP technologies on the workstations would then allow the project team to communicate even more effectively. In fact, I am quite surprised VOIP isn't integrated into Gnome at all without 3rd party tools. Just just seems to be the next logical step. All of the sudden, video-, voice- and text-based communication become a big time reality.

File Sharing: Heck, while we're at it, why don't we add file sharing to the mix? I mean honestly, why isn't there a simple application that takes advantage of this simple idea? Instant messaging clients got it right, but for some reason, desktops haven't even started thinking about it. I guess this opens up a big hole on your workstation, so rules and featuresets would definitely have to be implemented, however, it is silly to pull up and email attachment, your USB thumb drive or an IM client just to share files. They all post the same risk anyhow.

Secure Shell GUI Functions: Why, when I pull up Konqueror, can I just type in sftp://, and have a GUI interface to a SSH account- icons and folders to all my files on the remote machine? Because it wasn't difficult to create and implement, and it just makes sense. Naultilus should be able to pull the same stunt without bloating the interface too badly. I shouldn't have to download 3rd party tools to do the same, although there are many 3rd party tools that have better features and functionality. Sure, sftp:// isn't a standard protocol, and just one the KDE team developed, but it is open source, so I am sure the Gnome devs can copy the code, and put it right into Nautilus. SSH in a terminal is cool, but SSH in a GUI is better.


More Command Line Tools: There are just some apps that make sense utilizing the command line. Again, Konqueror got it right by implementing a shell right into the GUI itslef. Pulling up such a shell ina Konqueror window gives me quick access to command-line tools that I wouldn't otherwise have without pulling up another application, and using more RAM and processes. Why have double the application, when a single app could handle it just fine? And such a simple implmentation too.

Menu Editing: Why does Gnome not have a builtin menu editor? I mean, this is just silly. Here, I use OSS, and I can't even have basic control over my desktop? I don't think much needs to be said here, except for right-clicking menu items, and either moving them to another menu or submenu, renaming them, chaning the properties of the menu item and removing it if necessary. Windows got this right. After all, it is my menu, why do I have to put up with what Gnome thinks is best for me?

Eye Candy, Eye Candy, Eye Candy: KDE has got you beat by a long shot with this. Infact, every other window manager does. Even Fluxbox. Where is the eye candy? Now, let me state that I am NOT an eye candy person. I prefer my RAM to look, however, I also do not represent the greater majority of Linux users. People want look, and they want it to look good. When, and where, has the GUI been heading, say, oh, the PAST 8 YEARS? As far as I can tell, no where. It still looks the same. Configuring this eye candy should be able to be handled just about anywhere. Sure, there would be a seperate "catch all" program that would handle themes, colors, icons, desktop wallpapers, window borders, styles, fonts, etc., however most every program should be able to take advantage of simple features such as transparency, window snapping, and so on. In other words, the GUI needs a complete overhaul, and the API needs to be extended to any GTK+ installed application for basic eye candy control. And don't tell me Gnome can't, because it doesn't want to get bloated and sluggish. I have news for you, it already is bloated and sluggish. Again, KDE on Ubuntu is faster and slimmer than Gnome on Ubuntu.


I know I only covered two major topics for more tightly interating the Gnome desktop, but they are the biggest in my mind. Again, if these features are already implemented, please, point them out. I am not a Gnome expert, and as such, may have overlooked something. I am just concerned about direction. It seems to me that Gnome seems to be lacking it. Rather, the Gnome developers seem to be tied down with trivial feature requests and bug fixes. I say step it up. Take some risks. KDE is the preferred desktop, because Gnome is losing users. Gnome is losing users, because of these issues, or lack thereof. Even Linus himself said "use KDE".

I am certainly open to criticism and ideas about this. By all means, either send me a trackback or comment below. I think it would be healthy to stimulate good conversation about the direction of Gnome and whether or not integrating applications into Gnome is a good idea.

{ 1 } Trackback

  1. Johns Blog » GNOME on the horizon | April 26, 2006 at 2:37 am | Permalink

    [...] My responses to this well composed post of GNOME constructive criticism. [...]

Post a Comment

Your email is never published nor shared.