Slashdot Mirror
Ask Slashdot: How Would You Fix the Linux Desktop?
itwbennett writes "Slashdot readers are familiar with the Torvalds/de Icaza slugfest over 'the lack of development in Linux desktop initiatives.' The problem with the Linux desktop boils down to this: We need more applications, and that means making it easier for developers to build them, says Brian Proffitt. 'It's easy to point at solutions like the Linux Standard Base, but that dog won't hunt, possibly because it's not in the commercial vendors' interests to create true cross-distro compatibility. United Linux or a similar consortium probably won't work, for the same reasons,' says Proffitt. So, we put it to the Slashdot community: How would you fix the Linux desktop?"
And that right there is the problem.
Works on My Machine.
There are so many different configurations for computers and new and emerging tech, and the testing and documentation so spotty, that you've got to run through dozens of websites to get your computer to work. It took me a YEAR to get support for an Elan touchpad. Someone else decided that the ath9k driver should fill with a random number after sleep or hibernation. What the fuck is wrong with that person? Oh sure, I could fix it by bringing up a window, rmmod / modprobe ath9k, but that was seriously every time I closed the lid.
Other problems were solved with one of the following:
"LOL get a new computer."
"It's not a problem with this part, it's a problem with THIS part. Report it to them."
"Sorry, my part is perfect, so you must be a crazy person. You could try this patch though."
YOU ASSHOLE I JUST WANT TO CHECK FACEBOOK NOT RECOMPILE A FUCKING OS.
And I'm not a slouch here, the post where you figure out how to add my particular computer to the specific commands to allow Fn functionality was mine. (Someone else did the heavy lifting, I put the last pieces together.)
So what would you do to fix it? The easiest thing to do would be check the hardware during the install process or as part of the Live CD. "This touchpad is giving a weird answer to the magic knock, support may be limited."
Then actually allow for easy tweaks to the UI. How do you change the login screen? What about sounds? Your average user wants to be able to do this. It's a motherfucking nightmare to do this in the Super-Friendly distro.
If you have to get anyone anywhere to press CTRL-ALT-T to install a repository, then you've fucked it up. End of story.
---
ECHELON is a government program to find words like bomb, jihad, plutonium, assassinate, and anarchy.
I can still recall when it was described as being the graphical environment for GNU software.... lost a lot of interest when that went away.
William
Sphinx of black quartz, judge my vow.
The lack of Visual Studios on Linux isn't something that is wrong with Linux so much as it is something that is wrong with Visual Studios. But I think it is definitely a valid point that the lack of some specific software is a deal breaker for many people.
For me, Adobe Creative Suite is the main thing stopping me from making the switch to Linux permanently, and I imagine I'm not alone. Sorry, GIMP is just not a replacement for Photoshop/Fireworks/Illustrator/Acrobat/Flash and the other dozen programs of varying usefulness.
The chance that Microsoft would ever consider porting Visual Studios is probably zero, and even then I imagine it's probably just not possible. I'm not holding my breath for Adobe either, but I have a small amount of hope that someday they may see the light.
The main roadblock is that the market has been dominated by a single vendor since long before a single line of the Linux kernel was written.
Actually the key to Microsoft's success was exactly the opposite of what you said. They were not the vendor, at least not to the general public. It was the likes of IBM, Compaq, and Dell who sold computers that ran DOS. It was the fact that there were multiple vendors that drove down prices to make the PC compatible the affordable solution with the widest selection of software.
IBM tried unsucessfully to stop the clone market. Apple too had clones, but they ended up more successful at eliminating their competition - and they ended up with a pitiful market share to show for it. By the time Apple started its official clone program to expand the Macintosh market share, it was too late.
This dominant vendor was nearly able to kill off Apple with an OS that has no GUI and required MANUAL MEMORY MANAGEMENT.
I think you are forgetting about the manual memory management of the original Mac OS. You had to specify how much memory each program would use. You are right about the lack of GUI. This is one of those times where being the better does not equal success. Being cheaper can often be more important. Of course, by that reasoning Linux should takes over!
Some people are under the delusion that magically turning Linux into a Windows clone or a MacOS clone would help anything.
Absolutely. If you attempt to look like someone else's product then people will only notice the differences as being inferior to the original. Linux needs to keep its own identity.
And that is why it will never grow to become the new standard OS: there is not one single indentity. I said before that DOS succeeded because it ran on computers from multiple companies. If you turned on a PC compatible then you faced a familiar interface: C> This is not the case with Linux. Every distro does things their own way, with different windows managers and methods of installing software. It will never gain mainstream status while it appears to me hundreds of different operating systems.
But it is this flexibility and configurability that is what is so good about Linux. If you search long enough, you will eventually find the distro that fits your needs. And if you can't find one, you can make your own. But this means that it will always be a niche market for people who like to tinker and experiment with their computers.
So as others here have said, there is nothing wrong with the Linux desktop that needs to be fixed. All that needs to change is our expectations that the general population is ready and willing to put the time and effort into adapting to the OSS way of doing things when what they use now does everything they need.
I use LXDE because it is light on the resources. I really do not see why a desktop environment has to use several hundred megabytes of RAM. But even LXDE takes a lot. Needs about 100M. It makes them sluggish. The Firefox developers embarked on a "memshrink" program, and it's yielded excellent dividends. That effort made Firefox faster and more reliable. Seems Linux desktops could benefit from a similar hard look at memory usage.
LXDE has other problems. The file manager, pcmanfm, is still buggy and prone to crashes. Move lots of files around with it, and its stability goes to pot. It'll quit handling commands when it doesn't just crash. I've had to close it and start it up again to get it to work properly. I've not had good experiences with KDE or Gnome's file managers either. The file manager is a core part of any desktop environment, and the ones available in Linux are not good enough. Then there's the window manager, Openbox. Openbox works fine, but it isn't easy to configure. It has unusual commands (Shade/Roll up/down, and Un/decorate) that only serve to confuse the casual user, and which cannot be removed. If I switch to, say jwm, which doesn't have such extraneous features, then I have to deal with lxpanel and jwm's dueling task bars.
The UIs of all these desktop environments are full of holes and missing functionality. Still difficult to do it all and not at some point drag out the old text editor. For an example of a hole in the functionality, in LXDE if you right click on the desktop, a window pops up. Fine so far. Then if you click on the main menu (aka Start) button, that popup window does not go away, and the menu does not come up. You have to click somewhere on the desktop to make that popup window go away, then you can access the menu by clicking on the magic button. Why does it work that way? It's kludgy, that's why. The Linux desktop is still a messy collection of independent apps that don't play nice with each other. It lacks polish.
Peripherals are another weak spot. What happens if you try to print something, but you forgot to turn the printer on first? Depends how CUPS is configured. That job could hang around in the queue forever, and you will not be able to print until it is cleared. And it can't be cleared by any action that makes sense to a casual user. Canceling the job is the way to get printing working again, but this is not so easy. Turning the printer on doesn't work. Even rebooting doesn't work. But first, the user may not know any of this is happening, and will try to print again. Might end up with multiple copies. There may not be a printer dialog in which the user can cancel a job, instead the user has to pull up a browser and navigate to localhost:631. Or bring up ye olde command line prompt and do "lprm *". How many casual users know to do that? Evidently HAL was a wrong turn, and now it's all dbus.
One other thing: games. For games, must have hardware accelerated 3D graphics on commodity low end graphics cards. The open source drivers still can't do it. The proprietary drivers can, but cause other problems.
Intellectual Property is a monopolistic, selfish, and defective concept. It is "tyranny over the mind of man"