One more Gnome rant

I was thinking yesterday about Matthew Garrett who made a Guadec presentation about “Bluetooth in Gnome” and how it can become more complete. For once more though I could not stop thinking that without integration with the underlying system you can never have a GUI that works as the user expects it. The way Windows and OSX and BeOS work that is.

For example, you might have a Bluetooth/IrDA GUI to send files to another such device, but you don’t have a GUI to start/stop the Bluetooth service or to pair the two devices in the first place. Or, you might have gnome-power-manager, but you don’t have a GUI to really configure ACPI (g-p-m expects that the underlying scripts are already configured by the distro). Or that you might have a mouse preference panel, but you don’t have a way to configure the laptop’s touchpad properly. How many times we have asked for a feature in a Gnome app and the reply we got was “I can’t add this feature because it’s distro-specific“.

Some will take the easy way out and say: “this is the distribution’s job”. I beg to differ: Different implementations/scripts by different distros have created the incompatible forked chaos we have today in desktop Linux. LSB came in too late and it doesn’t cover all aspects. If Gnome was to have GUIs for more underlying works, it would force other distros to become more compatible. At the end, the user and the desktop will be the winner of the situation.

In essence, I would like to see Gnome fully supporting the internals of the 3 most popular Linux distros: with Ubuntu, SuSE and Fedora being the primary targets (and Debian and Mandriva being the underdogs). Please note that I don’t use any of these distros, I use Arch Linux instead. But I recognize the need for uniformity and access to the internals of the most popular systems. The desktop Linux will gain more followers if the GUI was more powerful and did what the user needed to do instead of having to resort to either the command line or the distro-specific preference panels (which sometimes don’t exist for specific operations especially in the Ubuntu’s case).

Please be very careful: I am not suggesting that Gnome (or KDE for that matter) rewrites all the system panels of these distros. This is NOT what I advocate. What I am suggesting though is that Gnome finds a way of “understanding” the architecture/scripts/philosophy of these few supported distros and so when there is a GUI Gnome app that needs to be written, it uses 100% the system in a way that gives power to the user the way he would have that power with OSX and Windows. For example, if a user wants to send a file via Bluetooth, Gnome should ask “Bluetooth is currently OFF. Start it? Yes/No” and have Gnome “understand” where the service is and how to start it (given that the permissions of the system allow this operation). You see, without such power within Gnome and its apps, having a dry “obex BT file exchange” utility only does HALF of the job for that user that recently came from Windows or OSX. Integration is what’s missing from the Linux desktop today.

Here is my train of thought: if Gnome could do the above, it would properly serve 90% of its desktop Linux users. If Gnome continues to try to be as OS/distro-non-specific as it is today, it will cover 100% of the userbase, but it will serve them in a mediocre way. If there are 10 million Gnome users today, it means that it screws 9 million users because of that other 1 million users who use the least popular OSes/distros (distros that will eventually add support for the new Gnome requirements anyway). It’s a hard decision to take, but this is the kind of decision that a true leader would be able to get. Unfortunately, Gnome is leaderless. Its Foundation can’t get such decisions and the maintainers don’t have such power.

And we, the users, are back to the command line, editing Xorg.conf to make our touchpad scroll. :P

Post a comment »

Lumbergh wrote on July 2nd, 2006 at 8:43 AM PST:

Of course that’s all true, but if you generalize what you said, I think you would come to the conclusion that the real problem is the concept of “distros” itself. The current model of desktop Linux (including distros) is a complete failure. Duplication of effort, dilution of talent, appealing to a lowest common denominator, and irrational ideological premises are why desktop linux is still stuck in low single digits, and why the current linux desktop paradigm has no hope of ever breaking out.


This is the admin speaking...
Eugenia wrote on July 2nd, 2006 at 8:50 AM PST:

I agree, but people want to make money out of the whole thing, so the creation fo the distros was inevitable. But I agree, when you see the whole desktop linux thing from further away, having distros fragments and slows down the adoption and fast evolution and maturity of the project.


Joseph wrote on July 2nd, 2006 at 8:56 AM PST:

>Of course that’s all true, but if you generalize what you said, I think you would come to the conclusion that the real problem is the concept of “distros” itself. The current model of desktop Linux (including distros) is a complete failure. Duplication of effort, dilution of talent, appealing to a lowest common denominator, and irrational ideological premises are why desktop linux is still stuck in low single digits, and why the current linux desktop paradigm has no hope of ever breaking out.

The concept of distros itself doesn’t harm anything, as long as they adhere to certain standards. Out of the millions of distros there are only a few which actually ‘matter’.

And it’s really such a wonderful thing. I relatively insignigant in the overall percentage of consumer desktop users.

That said, I hope Novell&SLED steal away a large amount of business from RedHat, or just gain it at their expense. I don’t mean to sound ungrateful, they have done a lot. I don’t mean to sound greedy when I say I want more — oh, wait. I do. They exist for the consumer and not the other way around, more desktop innovation please. Can’t live on the past forever.

As for the basic premise of the rant, I agree. It can easily be done independant of the GNOME community if it’s that useless. YAST goes a long way for SLED, maybe not far enough. Wish they was a GTK+ frontend though (even though I prefer Qt, it just stands out like a sore thumb)


Joseph wrote on July 2nd, 2006 at 9:04 AM PST:

It murdered my post? wth?

FIX:

And it’s really such a wonderful thing. I love Gentoo but 90% probably hate it. They can just go use Ubuntu, SuSE, or Fedora, the big three. Apple is proof you can have a superior product and still remain relatively insignigant in the overall percentage of consumer desktop users.


Lumbergh wrote on July 2nd, 2006 at 9:51 AM PST:

>I agree, but people want to make money out of the whole thing, so the creation fo the distros was inevitable.

Most people don’t make money off of distros. Even those that try to make money off of distros (apart from support/service plans) don’t really make money. Novell is finding out the hard way. Novell executives must curse Shuttleworth every morning they wake up;) It’s hard to make money when you’re trying to divide up an already meager pie and most people are already used to getting their pie for free.

But the real issue is, of course, who is competing against who. Does it make sense for Novell to be competing against RedHat, Microsoft, and Sun? Novell dropped the ball regarding Does it make sense for KDE and Gnome to be competing against each other?

Most distros are about the same, but different enough that ISVs have no target (something Portland Project is working on). That’s the real showstopper.

I think the best thing to happen in the short-term would be a Novell/RedHat merger – or anything that would help the fragmentation problem.


Lumbergh wrote on July 2nd, 2006 at 9:56 AM PST:

>Apple is proof you can have a superior product and still remain relatively insignigant in the overall percentage of consumer desktop users.

Apple’s problems are that most people don’t want to pay the Apple hardware tax, not enough hardware choice, XP is good enough, and windows has the most software.

Now that they’ve switched over to Intel, I would look for a partnership deal eventually.


nobu wrote on July 10th, 2006 at 6:02 AM PST:

Linux and all its flavors are great for the contributor. Those who use it, programmers, Network Admins, etc. all can and do contribute in some way. However, Windows and Macs are both consumer products and as a consumer they are not expected to contribute in any way shape or form.

The money that they paid for the OS and Hardware for them are contribution enough. In the linux world you do not have to contribute, but usually you do it because you would like something to work the way you wanted it and so you contribute. To make a Linux distro a consumer product will be quite difficult. If you focus too much on the consumer, the community of contributors, feeling betrayed will reject you. And their goes your Open Source advantage, example, Linspire.

I think right now the only distro that has a chance to cross over to the consumer is Ubuntu. They first started out with trying to build one of the best damn communities and are slowly working their way into different markets…..and they are doing it with community approval.

Red-Hat Novell even though they have innovated and come up with great open source software, they are so focus on the business side of it, they neglect their community. The feeling you get with ubuntu is that you are not outside looking in, you can see the inner workings and that I think makes all the difference.


Comments are closed as this blog post is now archived.

Lines, paragraphs break automatically. HTML allowed: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>

The URI to TrackBack this blog entry is this. And here is the RSS 2.0 for comments on this post.