Slashdot Mirror
Why Linux Is Not Yet Ready For the Desktop
An anonymous reader writes "Every now and then a new- or old-media journalist tries to explain to everyone why Linux is not yet ready for the desktop. However all those men who graduated from their engineering universities years ago have only superficial knowledge about operating systems and their inner works. An unknown author from Russia has decided to draw up a list of technical reasons and limitations hampering Linux domination on the desktop." Some of the gripes listed here really resonate with me, having just moved to an early version of Ubuntu 9.10 on my main testing-stuff laptop; it's frustrating especially that while many seemingly more esoteric things work perfectly, sound now works only in part, and even that partial success took some fiddling.
I've heard and read that mantra ten years ago. The future is not web-based because no large corporation will put/send/store their sensitive stuff (as in trade secrets) on any other corporation's web servers. Not even email. Ever.
It took almost 3 months to get the sound working on Ubuntu (TOS-link). Even to this day I'm scared that if I lose the system I'll lose the configuration- it required editing different accounts, adding new packages, modifying them in a non-standard fashion, adding options that weren't documented...
Windows XP? Put it in and the sound comes out.
I'll say the same thing about hard drives too- while the support is built in I still had to do some 20 commands to add, mount, locate, format, automount, edit the UUID manualy, fdisk....
Nothing better to kill 2 hours of your precious life.
Why Linux is not (yet) Ready for the Desktop
Preface:
In this document we only discuss Linux deficiencies while everyone should keep in mind that there are areas where Linux has excelled other OSes.
A primary target of this comparison is Windows OS.
Linux major shortcomings and problems:
0. Premise: proprietary software will stay indefinitely. Full stop. You may argue eternally, but complicated software like games, 3D applications, databases, CADs(Computer-aided Design), etc. which cost millions of dollars and years of man-hours to develop will never be open sourced. Software patents are about to stay forever.
1. No reliable sound system, no reliable unified software audio mixing, many (old or/and proprietary) applications still open audio output exclusively causing major user problems and headache.
1.1 Insanely difficult to set up volume levels, audio recording ... and in some situations even audio output.
1.2 Highly confusing, not self-explanatory mixer settings.
1.3 By default many distros do not set volume levels properly (no audio output/no sound recording).
2. X system:
2.1 No good stable standardized API for developing GUI applications (like Win32 API). Both GTK and Qt are very unstable and often break backwards compatibility.
2.2 Very slow GUI (except when being run with composite window managers on top of OpenGL).
2.3 Many GUI operations are not accelerated. No analogue of GDI or GDI+. Text antialiasing and other GUI operations are software rendered by GUI libraries (GTK->Cairo/QT->Xft).
2.4 Font rendering is implemented via high level GUI libraries, thus:
2.4.1 fontconfig fonts antialiasing settings cannot be applied on-the-fly.
2.4.2 Fonts antialiasing only works for certain GUI toolkits (see 2.1).
2.4.3 Default fonts (often) look ugly.
2.4.3.1 (Being resolved) By default most distros disable advanced fonts antialiasing.
2.4.3.2 By default most distros come without good or even compatible with Windows fonts.
2.5 No double buffering.
3. Problems stemming from the vast number of Linux distributives:
3.1 No unified configuration system for computer settings, devices and system services. E.g. distro A sets up networking using these utilities, outputting certain settings residing in certain file system locations, distro B sets up everything differently. This drives most users mad.
3.2 No unified installer across all distros. Consider RPM, deb, portage, tar.gz, sources, etc. It adds a cost for software development.
3.3 Many distros' repositories do not contain all available open source software. User should never be bothered with using ./configure && make && make installer. It should be possible to install any software by downloading a package and double clicking it (yes, like in Windows, but probably prompting for user/administrator password).
3.4 Applications development is a major PITA. Different distros can use a) different libraries versions b) different compiler flags c) different compilers. This leads to a number of problems raised to the third power.
4. It should be possible to configure everything via GUI which is still not a case for too many situations and operations.
5. Problems stemming from low linux popularity and open source nature:
5.1 Few software titles, inability to run familiar Windows software. (Some applications (which don't work in Wine) have zero Linux equivalents).
5.1.1 No equivalent of some hardcore Windows software like AutoCAD/3D Studio/Adobe Premier/Corel Painter/etc. Home and work users just won't bother installing Linux until they can work for real.
5.2 No games. Full stop. Cedega and Wine offer very incomplete support.
5.3 Incomplete or unstable drivers for some hardware. Problems setting up some hardware (like sound cards or TV tuners/Web Cameras).
5.3.1 A lot of WinPrinters do n
The vast majority of desktop computer users is happy with Minesweeper, Solitaire and Tetris.
The Tetris Company has never put out a product for Linux, except possibly the browser-based Tetris Friends. And it alleges that workalikes such as Lockjaw and Gnometris violate its copyright, though this US Copyright Office document makes Tetris's claims look flimsy.
"I would think that it was released they should have it mostly working well."
No, I would expect 9.04 to be mostly working well (which for me it almost does - the regression in the intel video card support is ticking me off though). 9.10 is at early alpha - I would expect it to not work very well at all. So the submitter's complaints about issues with 9.10 are unwarranted.
At some point, somewhere, the entire internet will be found to be illegal.
The future is web based.
Is it? After a typical month I am near my download limit for the month, and all it is is web browsing, email, and some file transfers. What is a web based solution going to do to bandwidth usage?
I've used Google docs for a quick project, and it has vastly cut and inflexible features compared to a spreadsheet installed on your machine.
Web based is too inflexible. Just my opinion of course.
Take Nobody's Word For It.
Not if it's an OEM license.
Mind you, I've used linux here and there since the 1.3 kernel (slackware then), and I've tried out just about every version of Ubuntu. This is the first time it stays in use.
Some things in TFA make me wonder though, like "Enterprise: no standard way of software distribution". How hard is it to set up a local repository(-ies), from where workstations get updates?
Finally, the next time someone posts and article about Linux and the desktop, please be clear which desktop we're talking about. This article seems to talk about all of them at once.
True confidence comes not from realising you are as good as your peers, but that your peers are as bad as you are.
Here in Europe, we have these technologies called 'GPRS' and '3G' which mean you're network connected over 95% of the land area.
We have that in the United States, but in this recession, not everybody who owns a laptop has 720 USD per year to blow on a 3G plan in addition to what they're paying for Internet access at home.
As an aside, the Intel driver thing was about to be a deal-breaker for me also, after two days of using 9.04. Then I thought there must be a way to load the 8.04 video drivers for it, and lo, there is!
Give that a try. I bet it fixes your problem; it worked awesome for me.
(I ran into an intractable network card issue with 9.04 though, which forced me to go back to 8.04 entirely, but at least this solved my video problem...)
mirrorshades radio -- darkwave, industrial, futurepop, ebm.
I strongly recommend you try Wine Doors if you haven't already.
It's probably not included in the default installation because I think you have to have a Windows license to install some of the DLLs and such (then again, who doesn't have a couple of those sitting around?)
OK, so let's deconstruct this point by point. I've left one or two points out where I have no specific comments.
0. Premise: proprietary software will stay indefinitely. Full stop. You may argue eternally,
but complicated software like games, 3D applications, databases, CADs(Computer-aided Design),
etc. which cost millions of dollars and years of man-hours to develop will never be open sourced.
Software patents are about to stay forever.
Bold predictions indeed. True, I think proprietary software will remain, particularly in the vertical market; however a certain segment of software will become commoditised (arguably some of it already has been) and therefore users will expect it to be free or priced lower than cost.
1. No reliable sound system, no reliable unified software audio mixing, many (old or/and proprietary) applications still open audio output exclusively causing major user problems and headache.
1.1 Insanely difficult to set up volume levels, audio recording ... and in some situations even audio output.
1.2 Highly confusing, not self-explanatory mixer settings.
1.3 By default many distros do not set volume levels properly (no audio output/no sound recording).
Couldn't agree more here. ALSA has improved audio in a few areas but in all other aspects, from a user perspective it has only made things more difficult. Someone else commented recently on Slashdot regarding the BSD approach to this problem, it sounds like they have done a lot better by staying with/improving OSS. I really wish someone would stand up and take charge of improving Linux's core audio infrastructure instead of putting band-aids like PulseAudio on top.
2.1 No good stable standardized API for developing GUI applications (like Win32 API). Both GTK and Qt are very unstable and often break backwards compatibility.
I'm not sure this is really as bad as is made out. In between major releases, Qt and Gtk both take backwards compatibility very seriously. Qt at least is a commercial product, they have a commitment to maintain compatibility.
2.2 Very slow GUI (except when being run with composite window managers on top of OpenGL).
Too general to respond to - can hardly be true for all machines.
2.3 Many GUI operations are not accelerated. No analogue of GDI or GDI+. Text antialiasing and other GUI operations are software rendered by GUI libraries (GTK->Cairo/QT->Xft).
I thought that was the point of Cairo... ? Not my area of expertise though.
2.5 No double buffering.
No explanation of how this is relevant to an end user.
3.1 No unified configuration system for computer settings, devices and system services. E.g. distro A sets up networking using these utilities, outputting certain settings residing in certain file system locations, distro B sets up everything differently. This drives most users mad.
Honestly I don't think the average user is really going to care where a configuration tool stores its settings as long as it works; only a power user or developer would. Of course it would be nice if people would use the same tools. However, although it's taken quite some time to work in all situations, NetworkManager has vastly improved network configuration ease of use and has been adopted by many distributions.
3.2 No unified installer across all distros. Consider RPM, deb, portage, tar.gz, sources, etc. It adds a cost for software development.
True, but arguably as far as the packaging alone is concerned, if you target RPM and deb you're going to cover most of the distributions that actually matter to end users.
3.3 Many distros' repositories do not contain all available open source software. User should never be bothered with using ./config
From my link
The problem is that at this point is been over a year and we have seen no progress. So it's hard to say if they are hard at work or gave up for now.
In Stockholm I pay $3-$9 per month for 3G, even with max data usage you wouldn't pay more than $360/year. Are you sure you're not using prices from 99?
As the A.C. below notes, not if it's an OEM issue of XP - the EULA ties XP to the machine it shipped with.
What he can do is blow away the XP install on that machine, install Linux and Virtualbox, then install that OEM XP as a Guest and do the activation over the phone. XP's EULA's are blissfully unaware of virtualizing, unlike Vista's which is, and does prohibit this trick.
Caveat: Some OEM releases of XP (looking at you HP) don't make nice - the installer looks for Vendor / Machine identifiers and will croak on ya. Dell plays nice, I did this with an 8600 / XP Pro rig.
I would add that when I tried installing Ubuntu a month or so ago on the same laptop, it said my wi-fi card was working, but it would not work. It also would not let me install the proprietary nVidia driver. When I ran the nVidia installer, it broke X.
You say you want a revolution....
My experience on this laptop (a Toshiba Equium M70-272):
;)
WinXP SP2 vs Ubuntu 7.04
Screen: default driver @ 800x600x16 vs default driver @ native resolution
Keyboard: default driver vs general driver
Sound: not recognised vs general driver
Wifi: not recognised vs Intel general driver
Printer: not recognised vs printer-specific CUPS support
Winmodem: not recognised vs default (non working) winmodem restricted driver
So... what is your point again?
For some reason, driver hunting for Windows is acceptable, but don't dare tell the guy trying Linux that Ubuntu might not pick up the play button on the side of the keyboard automagically!
When ideas fail, words become very handy.
Why would one who has windows license use linux?
They are masochistic?
There, fixed that for you.
Seriously, as much as I want to love Linux, and as much as I hate Microsoft, Windows gets the user experience down better than anybody except maybe MacOS. I've stopped using Linux on my personal laptop - installed because I got fed up with Vista's little quirks (should have rolled back to XP not Linux though) - in favor of my slower work laptop with WinXP on it because it just tends to work and I know how to do what I need it to do.
Linux is perfect for a home user as long as you meet a few criteria:
- You want to run servers but don't have the hundreds/thousands of dollars to lay down for a simple, effective solution (i.e. Microsoft).
- You are very technically inclined
- "Free" is significantly more important to you than "Easy" or "Simple".
- You are willing to put in the many, many hours you will need to learn the OS and how to configure it to do what you need
- You don't need or want any software or hardware that does not have a good Linux implementation/equivalent
There are others of course, but those are the big ones that come to mind for me. The last two in particular are why I can't use Linux. The first two are me, the third is true for me as well but not completely, and the last two don't fit me at all. I have a lot of better things I can be doing than spending hours trying to figure out why my sound doesn't work when I upgraded to a new version of my distro, discovering I have six different audio implimentations installed and only one of them will work.
Honestly, I could have copied my data and installed XP with less time and effort than it took to fix my sound in Linux, which broke after an update because things are not unified even in distros.
There is a reason *nix admins are few and highly paid in the server world, and it isn't because *nix is simple and easy. It's because *nix is very powerful for certain implimentations but it is notoriously difficult to manage. Until "notoriously difficult" becomes "easy" we won't see Linux on the desktop in any big scale. Ubuntu is better than any Linux I've used (and I've tried a bunch in the last 15 years), but it still doesn't touch XP, or really Win98 even. 95 it probably has beat though, heh.
Security is mostly a superstition... Avoiding danger is no safer in the long run than outright exposure. - Helen Keller