In response to the recent reviews of Debian 3.0 seen here and elsewhere, Debian developer Colin Walters has opted to start a Debian desktop subproject. The announcement e-mail states: “There is no reason we cannot make Debian easy to use for the masses. I personally want to be able to completely replace the Windows XP machine that my mom uses. We’re a fairly long way away from that, but I strongly believe in the power of free software. We’ve come this far, and we can go much farther.” The temporary location for the homepage of this project is here which will hopefully be moved to debian.org.
Debian Begins “Debian Desktop” Subproject
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
I think it’s about time Debian got something like this.
I’ve been using debian since early 1999 on my desktop as well as in the workplace (servers/workstations), so I’m no newbie, but I think for many things in debian, there is no reason at all not to make things easier, specifically with X, I don’t want to run some terse ncurses or text based program just to configure X!
I’ve been considering giving libranet a try though this gives give me hope in debian itself.
I’ve seen comments about Linux on Desktop is dead or something. The development of event since Open Office and mozilla.org 1.0 proved otherwise. Maybe it’s about time people take a look at what this little penguin is all about.
Good call, more users will be benefited from this I am sure.
This has been a long time coming, no we just need Slackware to shift some focus to the desktop
Why not just give your mom a dist besides Debian? There are others with a better desktop…
You may be a debian fanboy, but no need to make your mom one too.
I’ve been using Debian since “Buzz”.
It may not be the prettiest to install or the easiest to learn but I can make it “like I like it” better than any other distribution.
Debian will win out in the long run.
It has been my “desktop” for several years now.
From their home page:
Recognizing that both GNOME and KDE exist, we will support the use of both, and make sure they work well on Debian.
We support all desktop options: GNOME and KDE!!
We recognize that there are only two classes of users: the novice, and the expert. We will do everything we can to make things very easy for the novice, while allowing the expert to tweak things if they like.
We support all type of users: novices and experts!!
Other than that, it sounds like a good idea.
“I personally want to be able to completely replace the Windows XP machine that my mom uses.”
Well, if you’re gonna do that, then you better make sure that Debian can run all the software that Windows XP runs and supports all of the hardware that Windows XP does. Otherwise, unless you can convince the masses to care about your revolution (highly unlikely that’s ever going to happen), you don’t stand a chance.
Darius,
Windows XP cannot run all the software my linux computer can. Does that make it any less of a desktop competitor? To replace functionality does not have to mean running the same programs, just equivalents.
Nice to have a “free” desktop linux dist. I hope they don’t “integrate” the user-friendly-ness into main debian though, rather packages it seperately or something. Else they have one more moving to Gentoo here…
“Darius,
Windows XP cannot run all the software my linux computer can. Does that make it any less of a desktop competitor?”
As far as Joe User is concerned, since we live in a Windows-centric world, yes. It’s like the Betamax vs. VHS thing – one is of better quality, but the other rules the roost (DVD not withstanding, but you know what I mean).
“To replace functionality does not have to mean running the same programs, just equivalents.”
Yes, assuming that the equivalents have the same level of functionality, and that equivalents exist at all. For starters, find me an equivalent to this program (http://www.microsoft.com/streets) that I can run on my Linux box. (And if you dare say Mapquest, you have just given a prime example of why Linux isn’t ready just yet. The latter is much, much less functional, and only a trade-off someone would be willing to make if they were into the whole ‘revolution’ aspect of Linux.)
No, I don’t know of any equivalent to MS Streets, maybe RoadMap but since I’m not into this kind of programs I can’t really tell.
On the other hand this is not a typical joe user app, and there’s lots of non-typical joe user apps for linux that have no equivalents for windows as well. In addition to this there’s also alot of good apps for windows that are too expensive for joe user, while linux equivalents might be free.
If that’s not enough, many windows apps can be run in linux with wine.
In any case I still don’t belive Linux is the utlimate solution for the home desktops. It might be able to battle MS, and maybe even win one day. Still I don’t think a traditional unix system like linux is the best choice for a desktop. Time will tell though…
“No, I don’t know of any equivalent to MS Streets, maybe RoadMap but since I’m not into this kind of programs I can’t really tell.”
Based on the screenshots, the maps aren’t nearly as good of quality, and it lacks any sort of routing features as well. This is what you call one of those ‘drop in replacements’ that doesn’t quite do the trick – there seems to be quite a number of those in Linux.
“On the other hand this is not a typical joe user app”
Eh, I’m not so sure of how ‘un-typical’ it is anymore since they started including it with MS Works (which ships with a lot of computers now days).
But let’s assume (for the sake of arguement) that it’s not a typical Joe User app. Let’s say Joe User goes to his friend’s house and sees Streets & Trips running on his neighbor’s computer. “Wow, I’d like to run that!”, he thinks, but he can’t, because he’s running Linux.
See, the thing with people advocating Linux on the desktop is that they think the only apps Joe User will use are the ones that are available with the distro and with the ‘click and run’ interface. But what happens when Joe User wants to step out of the ‘sandbox’? Think he’s going to take the time to learn how RPM works? Dream on.
Go to any CompUSA (or whatever), stroll down the isles, and look at all the $20-$40 apps that are available .. the 3D Home Architechs, the Create-a-Cards, the Printshops, the Encartas, the educational software (Math/Science Advantage/etc), the games (Madden 2003 anyway?), the software that lets you make designs to iron on t-shirts, and all that stuff. Does Linux have equivalents for all of those? And if so, are they available within the ‘sandbox’ that Joe User is going to be trapped in?
Yes, you could argue that none of these are typical Joe User apps, but I would argue that EVERY Joe User is going to want at least one of these.
“If that’s not enough, many windows apps can be run in linux with wine.”
Yeah, and who’s going to set up Wine for Joe User? Though the Xandrose distro with Code Weavers might be a workable solution, CodeWeavers Wine does have its own sandbox and you still don’t have complete freedom to run what you want in the Windows world.
And this brings me to my last point … how many things in this world (that Joe User cares about) are going to be Windows only? Such as, Joe User buys a new portable MP3 player that interfaces with his computer. What are the chances that a Linux driver (or application) is going to be included with the CD? How hard would it be to find apps and drivers within the sandbox? If he buys a new scanner, are the ‘easy-as-pie’ buttons on the front panel going to work out of the box in Linux?
And what if it’s not a PC device at all? For example, a friend of mine recently bought his little girl a toy mouse that included software so you could hook it to a computer and ‘feed’ it? What do you suppose the chances are that thing is going to work on Linux?
No, I’m not here to say that one solution is better than the other on an even playing field. But what I am saying is that if Linux wants to appeal to the masses, it MUST be able to deal with just abut anything Joe User will run across in the Windows-centric world.
Either that, or else somebody is gonna have to change the world
First of all, I’m a Debian user and CS student, and while I love GNU/Linux and use Debian exclusively, I remain skeptic of its value as a “end-user” Windows replacement. My opinion is that Unix should be used for what it is good: servers and geeks.
That said, the project’s page is already at
http://www.debian.org/devel/debian-desktop/ .
Who’s Joe User ?
“Who’s Joe User ?”
Go to any Best Buy and you’ll see a guy there who’s looking for his first (or maybe second) computer amoung the HPs and Compaqs … that’s who Joe User is.
“Joe User” is NOT a person who knows of even thinks they know how to install an OS, including windows.
“Joe User” is a person who knows how to check email and use Microsoft Word. If they know how to get online, most likely AOL, that is a plus.
Please no flaming, but it is the truth.
If they ever manage to complete them – they still wouldn’t be ready for the desktop. Vanilla GNOME and KDE are quite confusing. Who wouldn’t have guess: they are made by UNIX geeks. It has a lot of jargon, and integrating it into Debian wouldn’t really make it easier to use – just more accessible to the user.
Recognizing that both GNOME and KDE exist, we will support the use of both, and make sure they work well on Debian.
I would say pick one and dump the other. If you like GNOME, push it to the limit, while give KDE the cold shoulder. Do the vice-versa if you like KDE. Or better still, have two teams – one on KDE, the other on GNOME. Combine both works.
What should be done is when a user pick, say KDE, everything would work the KDE way. When he wants to configure his machine, all the Debian config stuff would be in KControl. Same way for GNOME.
We recognize that there are only two classes of users: the novice, and the expert.
Then take the novice user and split them into several categories. Home users, corporate users, Rich home users with expensive gadgets like digicams and dvcams, gamers, etc.
I would highly recommend splitting the project to meet the goals of these different target markets, instead of just bundling them in one group. You can’t please them all. Sure – merge them up into one product to be release.
We will try to ensure that software is configured for the most common desktop use. For instance, the regular user account added by default during installation should have permission to play audio and video, print, and manage the system through sudo.
And they shouldn’t ever hear the word “sudo”.
We will integrate an easy-to-use system management utility into the standard Debian installation
I’ve got a better idea. Integrate them into KDE and GNOME. For example, say a person wants to configure a home network. He shouldn’t have to decide which control panel he should look. He should go straight to KControl.
Now, I have better goals for you.
Make something where the user don’t have to hunt a manual to use. Each time you ask a question, present the information in a less frightening way. If there is choices, clearly, in English, explain the choices. Don’t ask questions novices would never care, like “Which would you like. LiLo or Grub?”
For the desktops, as I said earlier, really integrate Debian into it. Make both of them easier to use for the novice. Yet not annoying for the expert. Make sure the expert don’t have to screeeammmm “WHY MUST THEY HIDE THAT”. Dumbing down a process is bad. Sure, hide something redundant from the novice, but always allow the expert to change it. Without reading the manual.
Finally, have your own style. Like Red Hat having Bluecurve. Avoid making something that looks like a rip-off of Aqua or Luna. Make something unique. Something unique enough that if someone walks into a room and looks at the screen, he can recognize that this is Debian – Not Mac OS X, not Windows XP, not some plain ol’ distro.
As I said easier – avoid dumbing down the process. Allow Experts to tweak their hearts content. Some distros instantly hide some parts of Linux just to make it less confusing to the user. Don’t do that. Explain to the user, in plain English what it is when he encounters it.
As for jargon – avoid it at all cost. For example, I think “Administrator” is a much more better, and more descriptive word, that “root”.
Now, the file hiearchy. The user should never ever see it. Hide it ala Finder in Mac OS X. Present something easier to browse. For experts, allow them to quickly revert back to the original on.
Then start using other distros, like Red Hat, Xandros, etc. waste some money. Take out a notepad, and list down what you really like about the distro that a novice would like. I like the way Red Hat mounts optical drives. It appears on the desktop, and an application would pop up for you to use it. I like the way Mandrake handles digital cameras by mounting them on the desktop.
Then really see what is wrong with them all and fix them.
And one last point: don’t clone Windows. It isn’t the best UI out there. Your grandma probably never really know how to use it. Make something better.
If they ever manage to complete them – they still wouldn’t be ready for the desktop. Vanilla GNOME and KDE are quite confusing. Who wouldn’t have guess: they are made by UNIX geeks. It has a lot of jargon, and integrating it into Debian wouldn’t really make it easier to use – just more accessible to the user.
Recognizing that both GNOME and KDE exist, we will support the use of both, and make sure they work well on Debian.
I would say pick one and dump the other. If you like GNOME, push it to the limit, while give KDE the cold shoulder. Do the vice-versa if you like KDE. Or better still, have two teams – one on KDE, the other on GNOME. Combine both works.
What should be done is when a user pick, say KDE, everything would work the KDE way. When he wants to configure his machine, all the Debian config stuff would be in KControl. Same way for GNOME.
We recognize that there are only two classes of users: the novice, and the expert.
Then take the novice user and split them into several categories. Home users, corporate users, Rich home users with expensive gadgets like digicams and dvcams, gamers, etc.
I would highly recommend splitting the project to meet the goals of these different target markets, instead of just bundling them in one group. You can’t please them all. Sure – merge them up into one product to be release.
We will try to ensure that software is configured for the most common desktop use. For instance, the regular user account added by default during installation should have permission to play audio and video, print, and manage the system through sudo.
And they shouldn’t ever hear the word “sudo”.
We will integrate an easy-to-use system management utility into the standard Debian installation
I’ve got a better idea. Integrate them into KDE and GNOME. For example, say a person wants to configure a home network. He shouldn’t have to decide which control panel he should look. He should go straight to KControl.
Now, I have better goals for you.
Make something where the user don’t have to hunt a manual to use. Each time you ask a question, present the information in a less frightening way. If there is choices, clearly, in English, explain the choices. Don’t ask questions novices would never care, like “Which would you like. LiLo or Grub?”
For the desktops, as I said earlier, really integrate Debian into it. Make both of them easier to use for the novice. Yet not annoying for the expert. Make sure the expert don’t have to screeeammmm “WHY MUST THEY HIDE THAT”. Dumbing down a process is bad. Sure, hide something redundant from the novice, but always allow the expert to change it. Without reading the manual.
Finally, have your own style. Like Red Hat having Bluecurve. Avoid making something that looks like a rip-off of Aqua or Luna. Make something unique. Something unique enough that if someone walks into a room and looks at the screen, he can recognize that this is Debian – Not Mac OS X, not Windows XP, not some plain ol’ distro.
As I said easier – avoid dumbing down the process. Allow Experts to tweak their hearts content. Some distros instantly hide some parts of Linux just to make it less confusing to the user. Don’t do that. Explain to the user, in plain English what it is when he encounters it.
As for jargon – avoid it at all cost. For example, I think “Administrator” is a much more better, and more descriptive word, that “root”.
Now, the file hiearchy. The user should never ever see it. Hide it ala Finder in Mac OS X. Present something easier to browse. For experts, allow them to quickly revert back to the original on.
Then start using other distros, like Red Hat, Xandros, etc. waste some money. Take out a notepad, and list down what you really like about the distro that a novice would like. I like the way Red Hat mounts optical drives. It appears on the desktop, and an application would pop up for you to use it. I like the way Mandrake handles digital cameras by mounting them on the desktop.
Then really see what is wrong with them all and fix them.
And one last point: don’t clone Windows. It isn’t the best UI out there. Your grandma probably never really know how to use it. Make something better.
it takes debian how many years to stabalise 2.2? now they are doing 3.0, so in about 5 years we can expect the next release, and now they want to spread there developers even thinner across the board, and add a desktop supersystem.
go debian.
thats probably what, another 2 years onto the release cycle.
The long relese cycles haven’t been because of developers, rather politics
Now they have changed them though, to provide quicker releases and next stable has already been in testing for a while…
This is what you call one of those ‘drop in replacements’ that doesn’t quite do the trick – there seems to be quite a number of those in Linux.
Don’t put words in my mouth – I never called it a ‘drop in replacement’, I said “No, I don’t know of any equivalent to MS Streets, maybe RoadMap but since I’m not into this kind of programs I can’t really tell.”
I never claimed Linux is a good choice for the desktop, rather the opposite: “I don’t think a traditional unix system like linux is the best choice for a desktop”
Where Linux do have a possibility to do a short-term sucess on the desktop, is in the really budget market. Ie the Wal-Mart linux PCs shipping for $200. With free software this solution can be a blessing for people not being able to afford thousands just for applications.
A development community can only be strong in an area if they have serious core values that are advantageous in that area. A development community is only as good as the people in it.
>We’ve come this far, and we can go much farther.”
The classic linux desktop fallacy. The area linux has come so far in the server, not the desktop, and it has come that far because the linux development community has had core values that benefitted server development. Unfortunately, the linux development community has constantly refused to believe that the desktop is a radically different world from the server with radically different requirements and goals and a radically different definition of what failure really is. They have never had core values in the area of usability, and it is precisely this that has been the largest obstacle to an easy-to-use linux.
There’s no “getting there” when “getting there” means a serious attitude readjustment that most debian developers have tradionally and continuously refused to make. If the same debian developers who 5 years ago pooh-poohed graphical installers are the same ones who today are scrambling to create something easy enough for your mom to use, they’re really going to crash and burn at it. A leopard with bleached fur does not a lion make, nor is it any less without spots. A wolf in sheep’s clothing is still exactly a wolf.
Let’s face it… all desktop systems suck except those with no applications or users 😉 (BeOS, AmigaOS)
The release cycles can be considered a feature of Debian. Debian Unstable is just as stable as other “stable” distro releases of Linux. The Debian Stable release is very good, especially for servers, because package conflicts is extremely rare.
Besides, this in a way is good for the desktop. I can’t tell you how many reviews (real reviews, mind you) that complain about bugs that could be pointed back to the package (rather the code). If in 2-3 years they release Debian Desktop, I might recommend it (seeing how good it is in terms of ease of use, of course). Less problems. Sure – no bleeding egde software, but why would a novice need them?
Linux on the desktop had come very far. In fact, it you plop back 3-4 years ago, you would pull your hair screeaming “THIS CAN’T BE A DESKTOP!!!!”
For example, 3-4 years ago, graphical 5-6 step installers is a dream. Nobody have them. In fact, people laugh at the idea of having one. Now, almost all desktop distros have them. It is a big deal, 3-4 years ago, far less amount of people could install Linux. Now, anybody experience with computers could.
Same goes for other stuff, like the desktop enviroment. Go out, get KDE 1.x and compare it with KDE 3.x. And think to yourself how much the desktop have evolved.
Besides, this quote; “We’ve come this far, and we can go much farther.” was meant that Debian had come out from being a hobbist distro, a small one, to become a major distro. Nobody ever said Debian was good at the desktop. A bunch of Debian developers plan to change that.
ealm: Perhaps they are only good because nobody uses them and there isn’t any applications for it.
This is a bit off topic but… AmigaOS has lots of apps (50k+ IIRC) and BeOS truly is both easy to use, fast and have a good design and API. But that discussion is getting a bit boring by now